FEDERAL DEPOSIT INSURANCE CORPORATION

FDIC Community Banking Study December 2020

Table of Contents

Foreword...... I Acknowledgements ...... III Executive Summary...... V Chapter 1: Community Bank Financial Performance ...... 1-1 Chapter 2: Structural Change Among Community and Noncommunity Banks...... 2-1 Chapter 3: The Effects of Demographic Changes on Community Banks...... 3-1 Chapter 4: Notable Lending Strengths of Community Banks...... 4-1 Chapter 5: Regulatory Change and Community Banks ...... 5-1 Chapter 6: Technology in Community Banks...... 6-1 Bibliography...... i Appendix A: Study Definitions...... A-1 Appendix B: Selected Federal Agency Actions Affecting Community Banks, 2008–2019...... B-1

FDIC Community Banking Study ■ December 2020

Foreword

Eight years ago, coming out of the financial crisis, the throughout this country, which is why I made this update FDIC conducted a study of community banks. This study to the 2012 Community Banking Study a research priority was the first large-scale review of community banks ever in 2020. I instructed my research team not only to update conducted, and it recognized the importance of community key aspects of the prior study, but also to consider new banks and their unique role in the banking industry. As topics that are important to community banks, such a result of that study, the FDIC changed its approach to as regulatory change and technology. By continuing to identifying community banks. In general, community study community banks and providing that research banks are those that provide traditional banking services to the public—our stakeholders—we can continue to in their local communities. As of year-end 2019, there were identify ways that the FDIC can provide support to 4,750 community banks in the country with more than these institutions. 29,000 branches in communities from coast to coast. I would like to extend a special thanks to Diane Ellis, Since the 2012 study, community banks have proven to be Director of the FDIC Division of Insurance and Research, resilient. Relative to noncommunity banks, community for leading this effort. banks have had faster growth in return on assets ratios, higher net interest margins, stronger asset quality, I believe this work will provide continued recognition and higher loan growth rates. Community banks have of community banks’ strength, their unique role in the continued to demonstrate this strength during the banking industry, and their value to the public. COVID-19 pandemic. Jelena McWilliams The FDIC recognizes the role community banks play Chairman, FDIC in providing loan and deposit services to customers December 2020

FDIC Community Banking Study ■ December 2020 I

Acknowledgements

The FDIC Community Banking Study is a collaborative Doreen Eberley, Keith Ernst, Pamela Farwig, Edward effort to identify and explore issues and questions about Garnett, Robert Grohal, Don Hamm, Angela Herrboldt, community banks. The study was produced under the Travis Hill, Deborah Hodes, Alex LePore, Kim Lowry, leadership of Shayna M. Olesiuk of the FDIC’s Division Ashley Mihalik, Jonathan Miller, Patrick Mitchell, Sumaya of Insurance and Research. John M. Anderlik, Margaret Muraywid, Thomas Murray, Robert Oshinsky, Camille Hanrahan, and Benjamin Tikvina managed the study. Schmidt, Philip Shively, Michael Spencer, Aaron Wishart, and Katrice Yokley. The study’s primary authors are as follows: I am also grateful to FDIC staff and contractors who Chapter 1: Community Bank Financial Performance provided valuable research, analytical, editorial, technical, Christopher J. Raslavich and publication support, including Vaishali Ashtakala, Jeffrey Ayres, Jeffrey DeLuca, Olayemi Jegede, Alexander Chapter 2: Structural Change Among Community and Gilchrist, Alan Levy, Jane Lewin, Lynne Montgomery, Chris Noncommunity Banks Nardi, Daniel Nguyen, James K. Presley-Nelson, Shawn E. Eric C. Breitenstein Schreier, David Spanburg, Derek Thieme, Jose M. Torres, Chapter 3: The Effects of Demographic Changes on Donna Vogel, Daniel Weeks, and Kathy Zeidler. Community Banks Chester Polson The authors thank Brennan Zubrick, Kyle Zhong, and the Conference of State Bank Supervisors for permission to use Chapter 4: Notable Lending Strengths of survey data from the National Survey of Community Banks. Community Banks Robert M. DiChiara Finally, we are immensely grateful to the late Richard Margaret Hanrahan A. Brown, FDIC Chief Economist, for his leadership and Richard D. Cofer Jr. contributions to the study of community banking at the FDIC, including leadership of the foundational 2012 Chapter 5: Regulatory Change and Community Banks Community Banking Study, development of the community George French bank definition, and nearly a decade of research and analysis on the topic. Rich made lasting contributions Chapter 6: Technology in Community Banks to the FDIC mission and was in our thoughts as we put Daniel Hoople together this study. He will be missed.

I would like to thank the many senior officials and staff Diane Ellis, Director from the FDIC who provided valuable insights and feedback Division of Insurance and Research during the review of the study, including Leonard Chanin,

FDIC Community Banking Study ■ December 2020 III

Executive Summary

The 2020 Community Banking Study is an update to the During the period 2012–2019, community banks narrowed Federal Deposit Insurance Corporation’s (FDIC) first the earnings gap with noncommunity banks because of community banking study, published in 2012, and covers factors such as a wider net interest margin and stronger the period from year-end 2011 through year-end 2019. The credit quality. Community banks ended 2019 with a earlier work made several important contributions to our quarterly pretax ROA ratio of 1.44 percent, only 22 basis understanding of the performance of community banks points below the pretax ROA ratio of noncommunity banks, and the significant role they play in the banking system a significant improvement from the 43 basis point gap at and the nation’s economy. It also established a definition year-end 2012. Community banks maintained their margin of a “community bank” that was not solely driven by advantage by earning higher yields on earning assets, asset size but also incorporated a bank’s business plan, which was partly attributable to their holding a higher geographic footprint, and number of branches (Appendix share of longer-term assets than noncommunity banks. A). This study retains the definition established in the Community banks also maintained their asset quality earlier edition and updates several areas of analysis advantage over noncommunity banks as measured by including community bank financial performance, trends credit losses. The full-year net charge-off rate reported by in community bank consolidation, and community community banks reached a post-crisis low of 0.13 percent bank lending focus. The current study also extends the in 2019, which was 45 basis points below the rate reported conversation about community banks in several directions: by noncommunity banks. it broadens the analysis of demographic changes affecting community banks and the products and services they One area that noncommunity banks outperformed offer, and it provides both an analysis of the effect of community banks was noninterest expenses. regulatory changes on community banks and an account of Noncommunity banks were able to reduce overhead community banks’ adoption of new technologies. Finally, expenses from 3.01 percent of average assets as of year-end each chapter in this study concludes by suggesting—from 2012 to 2.61 percent of average assets as of year-end 2019. the perspective of the subject of the particular chapter— Community banks saw their overhead ratio decline from possible effects the COVID-19 pandemic could have on 3.13 percent to 2.83 percent during the same time period. community banks. Structural Change Among Community and Community Bank Financial Performance Noncommunity Banks

Community banks continued to report positive financial After the 2012 study the banking industry continued to performance, including improving pretax return on consolidate, but existing community banks were less assets (ROA) ratios, a wide net interest margin, and likely to close than noncommunity banks. Of the 6,802 strong asset quality indicators. Coming off the recession institutions identified as community banks at year-end that ended in 2009, community bank pretax ROA ratios 2011, just under 30 percent had closed by year-end 2019.1 steadily improved, increasing from 1.05 percent in 2012 In comparison, over the same period, more than 36 percent to 1.44 percent in 2019. The improvement in earnings of the 555 institutions that identified as noncommunity was widespread with over 60 percent of community banks had closed. Including institutions that were banks reporting increases from 2009 through 2019. community banks at year-end 2011 but noncommunity Community banks’ earnings performance, moreover, banks at year-end 2019 and vice versa, as well as banks improved relative to noncommunity banks. By certain newly chartered between 2012 and 2019, there were measures, particularly pretax ROA, community banks have 4,750 community banks and 427 noncommunity banks at long underperformed noncommunity banks. The most year-end 2019. important factor contributing to the earnings gap between community and noncommunity banks had been the ability of noncommunity banks to generate noninterest income— 1 The 2012 Community Banking Study reported 6,799 community primarily from investment activities that typically are not banks and 558 noncommunity banks . These numbers have changed part of the traditional community banking business model. slightly reflecting new and revised Call Report filings that caused designation changes .

FDIC Community Banking Study ■ December 2020 V The drivers of banking consolidation shifted after the 2012 population inflows—community banks grew quickly and study. In that study, we showed how consolidation between profitably and supported communities with commercial 1984 and 2011 for both community and noncommunity and industrial (C&I) and commercial real estate (CRE) loans banks was driven by failures and charter consolidation. to help areas continue to grow. Areas with net outflows, Between 2011 and 2019, a period of economic recovery, on the other hand, appear to experience demographic and failures declined substantially, voluntary mergers between economic headwinds, causing banks in those counties unaffiliated institutions increased and became the to grow more slowly and have lower commercial lending predominant cause of the decline in the number of insured portfolios—conditions that could weigh on community depository institutions, and mergers between institutions banks in those areas. These demographic trends could also that were part of the same holding company fell. The result in greater consolidation in the future. major contributor to the overall decline, however, was the historically low number of newly chartered institutions: Notable Lending Strengths of between 1985 and 2011, 183 new institutions were chartered Community Banks per year on average, compared with four per year between 2012 and 2019. This combination of factors pushed up the Community banks by count represent the vast majority rate of net consolidation for the banking industry between of banks in the United States. By other size measures, 2012 and 2019 to 4.3 percent, compared with its average of however, community banks represent a considerably 3.2 percent during the years 1985 to 2011. smaller share: in 2019, they had only 12 percent of total industry assets and 15 percent of total industry loans. Despite holding a small share of total loans, community The Effects of Demographic Changes on banks are a key provider of funding for many local Community Banks businesses, most importantly by making CRE loans, small The changing demographic makeup of the United business loans, and agricultural loans. States affects demand for community bank services: as demographics change, banks see changes in their CRE Lending client bases and in the demand for loans. Two major CRE loans provide opportunities for businesses to own demographic factors considered in this study are median commercial property, for housing within communities, age and net migration flows. A comparison with the and for the provision of retail and other services to community-bank industry as a whole shows that between metropolitan, micropolitan, and rural areas. Community 2011 and 2019, community banks that were headquartered banks are an important source of financing for CRE as in counties at one demographic extreme—counties with evidenced by these banks’ loan portfolios, the types of lower median ages and the highest levels of net migration properties they finance, and the myriad locations of the inflows—experienced faster asset and loan growth rates, properties financed. The share of CRE loans community were more profitable, and had larger shares of business banks hold (30 percent of the banking industry’s CRE loans. Such counties tended to be in metropolitan areas. At loans) is large relative to the banks’ representation in the the same time, community banks that were serving areas banking industry. CRE lending also is widely distributed, of the country at another demographic extreme—counties with almost all community banks holding at least some with higher median ages and the highest levels of net amount of CRE loans, and many holding substantial migration outflows—experienced fewer opportunities for portfolios. Community banks originate various types of growth. Such counties tended to be in rural areas. CRE loans: multifamily lending grew in the years between 2011 and 2019, and community banks are active lenders to Community banks headquartered in areas simultaneously a wide range of industries, including industrial, retail, and experiencing two distinct demographic trends nonetheless hotel industries. saw consolidation trends that were similar to trends in the industry as a whole. As a result, these counties’ share of In addition to lending across industry types, community the total banking industry headquarters remained stable. banks have been active CRE lenders across all sizes of markets. In 2019, community banks headquartered in rural In areas of the United States that were arguably most and small metropolitan areas held more than two-thirds of thriving—those with a younger population and net CRE loans held by all banks headquartered in those smaller

VI FDIC Community Banking Study ■ December 2020 geographic areas. In larger metropolitan areas, community Agriculture Lending banks’ share of loans was smaller, but still material. Community banks are an important source of financing Although community banks of all lending specialties for U.S. agriculture, funding roughly 31 percent of farm provide CRE financing, the share of community banks that sector debt in 2019, with half of that total financed by are CRE specialists increased during the period between community-bank agricultural specialists. The lending 2012 and 2019.2 These CRE specialists are important emphasis of community-bank agricultural specialists providers of CRE loans in small communities. largely played in their favor from 2004 through 2013. Community-bank agriculture specialists’ exposure to the As important providers of CRE financing, community negative credit effects of the housing crisis and recession banks are among those lenders interested in CRE market that followed was minimized, and instead they benefited dynamics in the years ahead. As of the beginning of 2020, from a strong, decade-long farming boom. Beginning the long economic expansion had been a positive backdrop in 2014, the agriculture sector struggled in terms of to CRE. Delinquency rates among community banks’ CRE profitability, but erosion in farm financial conditions loan portfolios had declined for much of the previous was gradual and generally modest. Credit quality at decade, and at the end of 2019, the average delinquency community-bank agricultural specialists weakened rate had settled at a very low level. but still remained favorable by long-term historical Small Business Lending comparison, and earnings and capital were strong.

Small businesses are key to the U.S. economy, representing Community-bank agricultural specialists tend to be the vast majority of all businesses by count and employing small; in 2019, more than 75 percent had total assets almost half of the private sector workforce. These under $250 million, and just 19 out of 928 community- businesses often need funding, for example for inventory, bank agricultural specialists had total assets in excess of working capital, or accounts receivable financing. $1 billion. Geographically, community-bank agricultural Despite holding only 15 percent of total industry loans in specialists were heavily concentrated in the center of the 2019, community banks held 36 percent of the banking country. Agriculture in this area is dominated by cattle, industry’s small business loans.3 Community banks focus corn, hogs, and soybeans and to a lesser extent cotton, on building relationships with small business owners dairy, poultry, and wheat. and tend to make loans that require more interaction with the borrower. By contrast, noncommunity banks, Community-bank agricultural specialists have shown a which dominate the smallest category of business loan strong commitment to lending to farmers through the originations—loans below $100,000 that are typically peaks and valleys of cycles in the agricultural sector. From business credit card lines—tend to use a scoring model first quarter 2000 through fourth quarter 2019, in only two that requires little interaction with customers.4 During quarters did community-bank agricultural specialists see the period covered by this study, community banks’ an annual decline in aggregate agricultural production loan share of small business loans per Call Report data has volume, and never in aggregate farmland-secured loans. declined. Small Business Administration 7(a) program loan originations increased from 38 percent of total Regulatory Change and Community Banks originations in 2012 to 46 percent in 2019 with many loans greater than $1 million originated. Finally, in response The period 2008–2019 was one of intense regulatory to the 2018 FDIC Small Business Lending Survey, many activity, much but not all of it in response to the 2008–2013 bankers said their C&I loans were extended predominantly financial and banking crises. So numerous were the to small businesses, supporting the widely held belief that regulatory changes that keeping current with them would many loans to small businesses are above the Call Report’s have challenged any bank, but especially a small bank with $1 million reporting limit. modest staff resources. While many factors affect banks and it is difficult to be definitive, the pace of regulatory change may have been one factor that contributed to three 2 Refer to Appendix A for specialty bank determination criteria . post-crisis developments: a high proportion (compared 3 This percentage is based on commercial and industrial and with other time periods and other banks) of small nonfarm, nonresidential loans below $1 million . community bank mortgage lenders that reduced their 4 Federal Reserve Banks .

FDIC Community Banking Study ■ December 2020 VII residential mortgage holdings, the record rates at which likely to increase their capital ratios in part by curtailing community banks exited the banking industry in the years loan growth. leading up to 2019, and an apparent increase in the target asset size of new small banks as reflected in their initial Finally, it is important to emphasize that this study equity capital. views regulations only through the lens of their effects on community banks; a discussion of the policy goals Based on their sheer number and scope, changes to rules Congress has sought to achieve with its statutes, or how regarding 1–4 family residential mortgage lending and well the regulations have achieved those goals, is beyond servicing have a strong claim to being the most important the scope of the analysis. Observations in this study about rules of the post-crisis period. Between July 2008 and the effects of rules on community banks should thus not be November 2019, largely in response to laws enacted to taken as criticisms of those rules. The overall thrust of the address abuses in subprime and alternative residential analysis, however, does support the idea that if the societal mortgage lending and mortgage servicing, federal agencies benefits of a thriving community banking sector are to be issued 36 distinct substantive final rules governing various preserved, it is important that regulations achieve their aspects of 1–4 family residential mortgage lending and public policy goals in ways that accommodate, to the extent mortgage servicing. Even so, community banks in the appropriate, the business models and learning curves of aggregate continued to grow their residential mortgage smaller institutions with limited compliance resources. portfolios. At the same time, noninterest expense ratios for community bank residential-mortgage lending specialists The chapter covers several types of rules beyond those increased relative to those ratios for other community mentioned here. Appendix B provides a chronology banks, and the proportion of community banks with and a brief description of selected federal rules and small mortgage programs that materially reduced their programs—157 of them—that applied to community banks mortgage holdings continued to increase. Both trends are and were put in place from late December 2007 to year-end optically consistent with the hypothesis that regulatory 2019 (an average of 1 every 28 days during the 2008–2019 changes affected the costs and level of participation period). in residential mortgage lending of some community banks. Developments in financial and information Technology in Community Banks technology also are likely creating a tendency towards commoditization of residential mortgage lending, with Community banks have adopted different technologies at effects on the distribution of mortgage lending across different rates, with newer technologies such as mobile banks of different sizes. Accordingly, it is not possible to banking, automated loan underwriting, and online loan be definitive about the relative importance of regulatory applications being no exception. According to research and changes in driving mortgage lending trends. community banks’ own descriptions of the opportunities and challenges, several factors may play an important role The most important change to capital adequacy regulation in community banks’ adoption of new technologies. These during the 2008–2019 period was U.S. implementation factors include a bank’s characteristics, the economic of a version of the Basel III capital framework. Leverage and competitive environment, and the attitudes and ratios of community banks increased faster and to higher expectations of bank leadership. levels than did those ratios for noncommunity banks, and their loan growth exceeded that of noncommunity banks Data from the 2019 survey conducted by the Conference of as well. A detailed look at how community banks brought State Bank Supervisors (CSBS) indicate that “low adopters” about the increase in their capital ratios shows that the of several recent technologies were distinguished mostly extent of asset quality problems played an important role by their smaller asset size and lower revenues. For at in influencing how banks responded. Specifically, healthy least some of the banks participating in the CSBS survey, community banks with low levels of nonperforming loans those same characteristics predated technology adoption, increased their capital ratios but do not appear to have suggesting that bank size and resources may indeed curtailed loan growth to do this, while community banks have influenced community banks’ decisions to adopt with higher levels of nonperforming assets were more technology. Although it is also plausible that the use of technology may have increased asset and revenue growth

VIII FDIC Community Banking Study ■ December 2020 after adoption, additional data and research are needed to presented to the economy, the banking industry, and determine whether that was the case. society at large cannot be overlooked. This uncertainty will present community banks with both challenges Community banks that had higher ratios of loans to assets, and possibilities. As earnings decline and credit losses higher growth, and better performance also were more materialize, community bank performance is likely to likely to have adopted the technologies covered by the deteriorate. The rate of community bank mergers may survey, even after differences in size were accounted for. initially slow but then rise as institutions reconsider Similarly, banks that faced greater competition, had more branching and location strategies. Changes in demographic optimistic expectations, and had more positive attitudes trends such as population migration away from urban toward technology were more likely to be “high adopters.” areas could benefit community banks located in more Certain factors were not associated with the adoption of rural areas by providing them with new opportunities technology or else made no difference that could not be for growth. At the same time, community banks that explained by asset size. Among these factors were loan specialize in certain types of lending that are centered specialization, deposits, location of main office, and local in metropolitan areas, such as C&I, could suffer with population. Future research into these relationships, increased loan losses or lower growth rates. Finally, as well as the methods community banks use to obtain the increase in demand for contactless ways of doing technology, will broaden our understanding of the key business may encourage community banks’ adoption of drivers, barriers, and risks associated with financial new technology or partnerships with financial technology technology and its likely effect on the continuing success providers. Overall, community banks have a strong history of community banking. of recognizing and meeting the needs of their customers, and community banks will continue this tradition in years Future Challenges and Opportunities for to come. Community Banks

Although our data for this study end with 2019, the significant uncertainty that the COVID-19 pandemic has

FDIC Community Banking Study ■ December 2020 IX

Chapter 1: Community Bank Financial Performance

Between year-end 2012 and year-end 2019, community banks with assets between $500 million and $1 billion banks continued to report positive financial performance, (1.44 percent). Smaller community banks (those with including steady improvement in pretax return on assets assets between $100 million and $500 million) reported a (ROA) ratios, a wide net interest margin, and strong asset pretax ROA of 1.41 percent while the smallest community quality indicators. While not the only way of measuring banks (those with assets below $100 million) reported a community bank performance, this analysis will address pretax ROA of 0.94 in 2019, which represents a 24 basis the comparison between these banks and noncommunity point increase from year-end 2012. banks. Community banks had long underperformed noncommunity banks, particularly in pretax ROA. Despite community banks’ positive earnings performance Although this trend continued during the years between between 2012 and 2019, their earnings were lower than the previous study and this one, community banks’ wider the levels reported by noncommunity banks—but the net interest margin and strong credit quality caused this difference narrowed: at year-end 2012 the gap was gap to narrow.1 The most important factor contributing 43 basis points, and by 2019, it had dropped to 22 basis to the earnings gap had historically been the ability of points. During the entire period, the average earnings noncommunity banks to generate noninterest income— gap between community and noncommunity banks in primarily from investment activities that are still not reported pretax ROA was 31 basis points. The trend in associated with the traditional community-banking reported pretax ROA suggests that community banks were business model. Additionally, in the years preceding 2019 able to manage profitability and could still effectively noncommunity banks’ noninterest expenses relative compete with their noncommunity bank counterparts. to assets had fallen below the comparable ratios of community banks. Nevertheless, by providing traditional Community Banks’ Net Interest Margins banking products and services to local communities, Exceeded Those of Noncommunity Banks for community banks remained profitable and were able Several Years to compete with their typically larger noncommunity bank competitors. Net interest margins (NIM) measure the spread between asset yields and funding costs for deposits and other borrowings. Wider NIMs result in higher levels of net Community Bank Pretax Earnings Increased interest income. In 2019 net interest income accounted Between Year-End 2012 and Year-End 2019 for over 78 percent of community bank net operating Coming off the recession that ended in 2009, community revenue. Community banks ended 2019 with a quarterly bank pretax ROA ratios steadily improved, increasing from NIM of 3.62 percent, exceeding the margin of 3.24 percent 1.05 percent in 2012 to 1.44 percent in 2019 (Table 1.1). The reported by noncommunity banks. Starting in 2012, improvement in earnings was widespread, with more community banks were reporting yields that were, on than 60 percent of community banks reporting increases average, 53 basis points higher than yields reported by throughout this study’s period. Larger community banks noncommunity banks (Chart 1.1). The primary way they (those with over $1 billion in assets) ended 2019 with the maintained their margin advantage was by earning higher highest pretax ROA (1.48 percent), followed by community yields on earning assets.

Table 1.1 Full-Year Pretax ROA (Percent)

2012 2013 2014 2015 2016 2017 2018 2019 All Banks 1.42 1.55 1.46 1.49 1.50 1.54 1.69 1.63 Community Banks 1 .05 1 .12 1 .19 1 .26 1 .30 1 .35 1 .42 1 .44 Noncommunity Banks 1 .48 1 .62 1 .50 1 .53 1 .53 1 .57 1 .73 1 .66 Source: FDIC .

1 A focus on pretax ROA, as opposed to return on assets after tax, facilitates comparisons between banks organized as C corporations, which are taxed at the bank level, and S corporations, in which tax obligations pass through to shareholders .

FDIC Community Banking Study ■ December 2020 1-1 Chart 1.1 Chart 1.2 Quarterly Average Net Interest Margin, 2009–2019 Quarterly Noninterest Income Percent Community Banks as a Percent of Average Assets, 2009–2019 4.0 Percent Community Banks Noncommunity Banks 2.5 3.8 Noncommunity Banks 3.6 2.0 3.4 1.5 3.2 3.0 1.0 2.8 0.5 2.6 2.4 0.0 2009 2010 2011 2012 2013 201420152016201720182019 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC. Source: FDIC.

Table 1.2 Assets With Maturities Greater Than 3 Years to Total Assets (Percent)

2012 2013 2014 2015 2016 2017 2018 2019 All Banks 28.8 29.5 30.2 31.6 32.4 32.6 32.4 33.3 Community Banks 42 .9 47 .3 47 .9 47 .4 47 .2 46 .8 45 .8 44 .8 Noncommunity Banks 26 .5 26 .7 27 .5 29 .2 30 .2 30 .5 30 .4 31 .8 Source: FDIC .

The higher yields reported by community banks were In 2019, noninterest income represented 0.9 percent of partly attributable to the fact that community banks held average assets at community banks, driving 20.2 percent of a higher share of longer-term assets, which typically their net operating revenue (Table 1.3). For noncommunity have higher returns than assets maturing in the short banks, noninterest income represented 1.5 percent term. As of year-end 2019, assets that matured or of average assets and drove 34.2 percent of their net repriced in more than three years accounted for over operating revenue. In 2019 noncommunity banks derived 44 percent of community banks’ total assets (Table 1.2). close to 18 percent of their noninterest income from The comparable figure for noncommunity banks was just market-sensitive revenue streams, including trading and 31.8 percent. Between 2012 and 2019 both community investment activities, which had not traditionally been and noncommunity banks increased their exposures to part of the community-banking business model. Instead, long-term assets because the historically low interest rate for additional income, community banks relied more environment had several effects including causing many heavily on asset sales and service charges, whether these banks to lengthen their balance sheets to help maintain activities were part of the bank’s strategy or not. their margins as well as meeting credit needs of borrowers looking to lock in at low rates for longer periods. Table 1.3 lists the categories of noninterest income reported by banks throughout the period of this study. Despite Compared with Noncommunity Banks, the granularity of these categories, many of the services Community Banks Generated Less offered by both community and noncommunity banks are Noninterest Income accounted for in the “all other” category—making it the largest component of noninterest income for both types Despite their net interest margin advantage, community of institutions. Banks are required to itemize amounts banks trailed noncommunity banks in overall earnings within the “all other” category only if the amounts exceed because of noncommunity banks’ ability to generate minimum levels. Thus, it is hard to compare with certainty higher volumes of noninterest income due to their business the relative importance of various components within the model (Chart 1.2). “all other” noninterest income category.

1-2 FDIC Community Banking Study ■ December 2020 Table 1.3 Noninterest Income at Community and Noncommunity Banks (Percent)

Full-Year 2012 Full-Year 2019 Category of Noninterest Income Community Noncommunity Community Noncommunity as a Percent of Total Noninterest Income Banks Banks Banks Banks Service Charges on Deposit Accounts 24 .3 12 .7 18 .8 13 .1 Fiduciary Income 6 .9 11 .9 8 .0 14 .3 Gains on Asset Sales 21 .7 3 .9 22 .0 4 .0 Market Sensitive Income1 2 .6 11 .8 3 .0 17 .6 Securitization Income 0 .5 0 .6 0 .1 0 .1 Servicing Income 3 .1 4 .7 3 .7 1 .2 Insurance Income 3 .3 1 .4 3 .1 1 .7 All Other Noninterest Income2 37 .5 53 .0 41 .4 47 .9 Total Noninterest Income 100.0 100.0 100.0 100.0 Noninterest Income as a Percent of Net Operating Revenue 22 .0 39 .4 20 .2 34 .2 Noninterest Income as a Percent of Average Assets 0 .95 1 .9 0 .87 1 .5 Source: FDIC . 1 Includes trading, venture capital, and investment banking income . 2 Other noninterest income includes service charges, commissions, and fees (such as safe deposit box rentals, money orders and cashiers checks, notarizing of documents, ATM fees, wire transfers), check sales, rental income from other real estate owned, bank- owned life insurance income, annual credit card fees and interchange fees .

Community Banks Reported Lower Levels Chart 1.3 of Noninterest Expense Relative to Quarterly Noninterest Expense Average Assets as a Percent of Average Assets, 2009–2019 Percent Community Banks 3.6 At year-end 2019, noninterest expenses at community Noncommunity Banks banks were 2.83 percent of average assets, down from 3.4 3.13 at year-end 2012 (Chart 1.3). The decline was due 3.2 primarily to reductions of premises and fixed asset 3.0 expenses as well as their “all other” noninterest expenses 2.8 (a category that includes items such as data processing 2.6 expenses, legal fees, and telecommunication expenses). 2.4 These reductions could be a combination of items 2.2 including reducing branches, streamlining computer 2.0 expenses, and lowering legal expenses. These declines 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 almost completely offset increases in salary and employee Source: FDIC. benefit expenses (Table 1.4). assets than noncommunity banks. Between 2012 and Historically (1987–2007) community banks reported 2019, however, community banks reported a noninterest lower noninterest expenses as a percentage of average expense ratio that was on average 18 basis points higher

Table 1.4 Compound Annual Growth Rate of Noninterest Expense Categories (Percent)

Full-Year 2012 Full-Year 2019 Community Noncommunity Community Noncommunity Banks Banks Banks Banks Salary and Employee Benefit Expenses 4 .6 7 .0 1 .4 3 .0 Premises and Fixed Asset Expenses 3 .2 5 .2 -0 .8 0 .7 Salary + Fixed Asset Expenses 4.3 6.6 1.0 2.6 All Other Noninterest Expenses 2 .4 4 .7 -1 .6 0 .8

Total Noninterest Expenses 3.8 7.4 0.1 1.9

Average Assets 4.0 9.3 1.5 4.0 Source: FDIC .

FDIC Community Banking Study ■ December 2020 1-3 than that of noncommunity banks. In the years leading up Chart 1.4 to the financial crisis, noncommunity banks had grown Quarterly Net Charge-O Rate, 2009–2019 their assets at a much faster rate than they had grown Percent Community Banks their noninterest expenses, which led to the convergence 3.5 Noncommunity Banks of the noninterest expense ratios of noncommunity 3.0 and community banks. After the financial crisis, 2.5 noncommunity banks continued to grow their assets at a 2.0 faster rate than their noninterest expenses, with the result that in 2019 they reported noninterest expenses relative to 1.5

2 average assets that were below community bank levels. 1.0

0.5

Community Banks Continued to Report 0.0 Low Levels of Credit Losses 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC. In 2019 the full-year net charge-off rate reported by community banks reached a post-crisis low of 0.13 percent. In general, between 2012 and 2019 loan portfolios

Community banks had generally reported lower loan- of community and noncommunity banks did not loss rates than noncommunity banks (Chart 1.4). This shift significantly. Community bank loan portfolios was especially true in the period 2008–2011, the years continued to remain heavily weighted toward nonfarm, during and immediately after the financial crisis. From nonresidential CRE loans and 1–4 family residential 2012 through 2019 community banks’ average loss mortgages. At year-end 2019, these two loan categories rates in commercial loan categories were comparable together represented 55 percent of community banks’ total to the rates for noncommunity banks (these categories loans, a share that had been relatively unchanged since include nonfarm, nonresidential CRE, and C&I loans), but 2012. Meanwhile, noncommunity bank loan portfolios community banks continued to report lower loss rates than continued to consist mostly of C&I loans, consumer loans, noncommunity banks in the two retail loan categories and 1–4 family mortgages—categories that, together, (residential real estate loans and consumer loans) as well represented over 62 percent of noncommunity banks’ as in agricultural loans (Table 1.5). total loans.

Table 1.5 Average Net Charge-Off Rate by Loan Type (Percent)

Loan Type Bank Type 2000–2007 2008–2011 2012–2019 Community Banks 0 .06 0 .53 0 .15 1–4 Family Noncommunity Banks 0 .13 1 .64 0 .26 Community Banks 0 .13 3 .68 0 .37 Construction & Development Noncommunity Banks 0 .12 4 .81 0 .23 Community Banks 0 .08 0 .55 0 .15 Nonfarm Nonresidential Noncommunity Banks 0 .10 0 .94 0 .13 Community Banks 0 .63 1 .41 0 .43 Commercial & Industrial Noncommunity Banks 0 .92 1 .56 0 .34 Community Banks 0 .87 1 .18 0 .85 Consumer Noncommunity Banks 2 .88 4 .81 2 .24 Community Banks 0 .12 0 .20 0 .08 Agricultural Noncommunity Banks 0 .27 0 .65 0 .20 Community Banks 0 .22 1 .02 0 .23 Total Loans Noncommunity Banks 0 .76 2 .24 0 .65 Source: FDIC .

2 Chapter 5 includes a more detailed analysis of community bank and noncommunity bank noninterest expenses .

1-4 FDIC Community Banking Study ■ December 2020 Summary Additionally, despite lowering their noninterest expenses, community banks were not able to match the strong asset Community bank performance between 2012 and 2019 growth that noncommunity banks had—growth that led to showed that despite lower earnings, community banks lower noninterest expense ratios for those banks in an area were profitable and could successfully compete against where community banks historically had had an advantage. their typically larger noncommunity bank competitors. Community banks continued to benefit from higher Overall, the performance of community banks continued margins due, in part, to their holding a higher share of to demonstrate that there is a role for these institutions long-term assets. Community banks also continued their in the banking landscape. Although they faced different long-run trend of strong asset quality metrics and lower challenges than noncommunity banks, community banks’ loan-loss rates. However, community banks continued proven advantages in the areas of net interest income and to lag their larger competitors in the ability to generate credit losses put these institutions in a good position to noninterest income, which appeared to be the biggest compete with noncommunity banks going forward. driver of the earnings difference between the two groups.

Box 1.1 Financial Performance and the COVID-19 Pandemic

Community banks faced significant challenges in 2020 amid the onset of the COVID-19 pandemic. Through the first half of 2020, community banks reported a decline in earnings primarily driven by significant increases in provisions for credit losses, as temporary shutdowns resulted in reduced business and consumer spending and uncertainty surrounding large parts of the economy continued (Chart 1.1.1). Community bank pretax ROA through the first two quarters of 2020 fell by 25 basis points to 1.19 percent from year-end 2019 as a result of the earnings decline. Comparatively, pretax ROA among noncommunity banks fell by over a full percentage point to 0.35 percent through the first half of 2020.

The NIM for community banks fell in the first half of 2020 as asset yields declined more rapidly than funding costs due to the low interest-rate environment and an increase in the volume of lower-yielding assets, including balances due from depository institutions and Paycheck Protection Program (PPP) loans. The NIM for community banks fell 11 basis points from year-end 2019 to 3.51 percent in the second quarter of 2020. Comparatively, the NIM for noncommunity banks declined 52 basis points from year-end 2019 to 2.72 percent, representing the lowest NIM for noncommunity banks on record.

Loan loss rates for community banks remained low and stable since the start of 2020. The net charge-off rate for community banks through the first half of the year stood at 0.12 percent, just 2 basis points above the rate recorded through the same point in 2019 and still well below the rate recorded by noncommunity banks. Minor credit deterioration among community banks has primarily been concentrated in the C&I, farmland, and agricultural production loan categories. However, community banks may report higher credit losses across other loan categories in the coming quarters as businesses and consumers continue to experience adverse effects as a result of government- mandated business and travel restrictions in response to the pandemic.

Loan growth served as a bright spot for community banks as loan volumes expanded at a rate that exceeded noncommunity banks in the first half of 2020. Community banks reported an annual loan growth rate of 13.5 percent in second quarter 2020. Comparatively, loan balances among noncommunity banks expanded by just 5.6 percent annually. The increase among community banks was driven by large increases in C&I loans and community bank’s participation in the PPP. As of second quarter 2020, community banks held 31 percent of PPP loans, with more than four out of five community banks (82 percent) participating in the program.

continued on page 1-6

FDIC Community Banking Study ■ December 2020 1-5 Box 1.1, continued from page 1-5

Chart 1.1.1

Contributors to the Year-Over-Year Change in Income for Community Banks First Half 2020 vs. First Half 2019 $ Billions Positive Factor Negative Factor 3.5 –$1.03 $1.42 $2.94 $2.08 $1.77 –$0.07 –$0.37 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 -1.0 –8% 4% 231% 24% 6% –17% –15% -1.5 Net Net Loan Loss Noninterest Noninterest Realized Income Income Interest Provisions Income Expense Gains on Taxes Source: FDIC. Income Securities

1-6 FDIC Community Banking Study ■ December 2020 Chapter 2: Structural Change Among Community and Noncommunity Banks

The decline in the number of banks that began in 1986 At the time the current study was being prepared, the continued through 2019. Between year-end 2011 and year- COVID-19 pandemic had not significantly affected the rate end 2019, the number of banks dropped from 7,357 to 5,177, of consolidation, although it ultimately may. Box 2.3 at the representing a decline of 30 percent. Among community end of this chapter contains an overview of the pandemic’s banks, the number fell from 6,802 to 4,750; among potential effects on consolidation. noncommunity banks, the number fell from 555 to 427. The Largest Components of Charter The drivers of net consolidation, however, shifted after Consolidation Between 2011 and 2019 2011. As described in the 2012 FDIC Community Banking Were Failures, Voluntary Mergers, and Study, a major cause of consolidation in the preceding New Charters two decades was bank failures, due mainly to the banking and thrift crises of the late 1980s and early 1990s and Charter consolidation is the sum total of failures, voluntary then to the financial crisis of 2007–2008 and the ensuing mergers, new charters, and other voluntary closings. Great Recession. But as the effects of the Great Recession subsided and the economy transitioned into a slow recovery Rates of Failure Declined followed by expansion, the number of failures declined. The merger booms of both the 1990s and the years following the Great Recession came close on the heels of Nevertheless, the average rate of net consolidation periods of economic and financial disruption, particularly continued to rise (Chart 2.1). The largest component of the disruption constituted by the banking crisis from consolidation identified in the 2012 Study—voluntary approximately 2008 through 2013. The financial crisis mergers between unaffiliated institutions—increased as had begun late in 2007, was quickly followed by the the economy recovered and expanded. At the same time, Great Recession, and roughly a year after the onset of the rate of mergers between institutions within a holding the financial crisis the number of bank failures began company declined. Finally, new bank charters became increasing (Chart 2.2). But in 2011 the failure rate started less common, meaning there were few new institutions declining, and by 2012 most of the failures associated with replacing those that merged, consolidated, or failed. the financial crisis and Great Recession had occurred.

Chart 2.1

Net Charter Consolidation Rates, 2009–2019 Annual Rates of Net Charter Consolidation as a Percent of Charters Reporting at Previous Year End Percent 0

-1

-2

-3

-4 Average: -5 2012–2019 –4.3%

-6 2009 2010 2011 2012 2013 201420152016201720182019 Source: FDIC. Note: Gray bar indicates recession period.

FDIC Community Banking Study ■ December 2020 2-1 Chart 2.2

Failure Rates, 2009–2019 Annual Rates of Failure as a Percent of Charters Reporting at Previous Year End Percent 0

Average: 2012–2019 –0.2% -1

-2

-3 2009 2010 2011 2012 2013 201420152016201720182019 Source: FDIC. Note: Gray bar indicates recession period.

Chart 2.3

Voluntary Closure Rates, 2009–2019 Annual Rates of Voluntary Closing as a Percent of Charters Reporting at Previous Year End Percent 0

-1

-2

-3

-4 Average: -5 2012–2019 –4.1% -6

-7 Voluntary Inter-Company Mergers Intra-Company Consolidation of Charters -8 Other Voluntary Closings -9 2009 2010 2011 2012 2013 201420152016201720182019 Source: FDIC. Note: Gray bar indicates recession period.

As the lingering effects of the recession wore off and 1990s (Chart 2.3). Many of the earlier mergers, however, economic expansion took hold, the failure rate continued particularly those occurring through 2000, were between to decline as failures became a much less important factor separately chartered institutions that were owned by the in charter consolidation. Between 2015 and 2019, only same holding company—that is, they were intra-company 25 institutions failed. In 2010, at the peak of the banking mergers.1 Starting in 2011, mergers were more likely to crisis, 157 banks failed. 1 The Riegle-Neal Interstate Banking and Branching Efficiency Act Voluntary Merger Rates Increased of 1994 removed many of the restrictions banks faced if they wished to open a branch in a different state than the one in which they Starting in 2011, rates of voluntary mergers rose to were headquartered . To the extent holding companies maintained separately chartered banks to comply with interstate banking levels not seen since the previous merger boom, in the restrictions, the Act rendered the separate charters unnecessary and facilitated their combination .

2-2 FDIC Community Banking Study ■ December 2020 occur between unaffiliated institutions—that is, they were In the meantime, mergers between charters within the inter-company mergers. same holding company dwindled as most had already consolidated their banks. A comparison of average merger Inter-company mergers reduce the number of genuinely rates for the two types of mergers shows that between independent institutions. Although intra-company 1985 and 2011, the unaffiliated merger rate averaged mergers reduce the number of chartered banks, because 2.3 percent of institutions per year, but 3.3 percent per year the merging banks are owned by the same holding in the period since 2011. In contrast, the intra-company company, such mergers can be thought of as combining merger rate averaged 1.5 percent per year between 1985 separate divisions of a single company rather than mergers and 2011, but only 0.6 percent per year between 2012 of distinct companies. and 2019.

Although between 2011 and 2019 unaffiliated (inter- A new type of voluntary merger occurred in 2012, when company) mergers constituted most merger activity for the first time a bank was acquired by a credit union. among insured institutions, the rate of such mergers did Between 2012 and 2019, 39 community banks were either not reach or exceed its previous peak. Between 1994 and acquired or were pending acquisition by 34 unique credit 1999 the annual average rate for inter-company mergers unions, compared with approximately 1,750 community was 3.6 percent, with a peak of 4.4 percent in 1998, banks that were acquired during this period by other but in the period after 2011, the annual merger rate for banks. For more information on the acquisition of unaffiliated institutions did not again reach 4.0 percent community banks by credit unions, see Box 2.1. until 2014, and it did not reach 4.1 percent until 2018.

Box 2.1 The Acquisition of Community Banks by Credit Unions Historically, credit unions and banks coexisted, offering similar services but with distinct business purposes. Although credit unions may have been viewed as competitors, they focused on a specific field of membership.a Mergers and acquisitions did not occur until 2012, when the first “purchase and assumption” of a bank by a federal credit union was completed.b

Credit unions continued to acquire banks after 2012, but the number of banks acquired by credit unions pales in comparison with the number of banks acquired by other banks over the same period. In the years since that first acquisition in 2012 through 2019, a total of 39 acquisitions of community banks by credit unions were completed or were pending.

Banks that were acquired by a credit union have some important characteristics that provide insight into possible reasons for their attractiveness to the credit union. Relative to otherwise similar non-acquired banks, acquired banks tended to be smaller in terms of asset size, have larger concentrations of single-family mortgage loans, and have smaller concentrations of C&I loans. These acquired banks also tended to have higher efficiency ratios and less profitability overall. Taken together, these characteristics suggest that the acquired banks were small enough that credit unions could incorporate the bank portfolio into existing operations. The banks also had loan portfolios that complemented the credit unions’ business models.

As of year-end 2019, the trend among some credit unions to acquire banks made up a very small portion of the overall number of banks acquired in mergers.

a Potential members must belong to a credit union’s field of membership in order to join . For example, membership in a credit union with a “community charter” is limited to people who live, work, worship, or attend school within a well-defined geographic area, such as a neighborhood, city, or rural district . Legislative and regulatory changes during the last 20 years, such as the Credit Union Membership Access Act of 1998, have increased the number of people eligible to join credit unions . b A purchase and assumption transaction involves the transfer of assets and deposit liabilities from one institution to another without the two institutions legally combining into a single entity . When a credit union “acquires” a bank, it purchases all, or substantially all, of the bank’s assets and assumes its liabilities . The legacy bank liquidates any remaining assets and relinquishes its charter . While credit unions had acquired assets from banks prior to 2012, there had not been a purchase and assumption of an entire bank by a federal credit union until then .

FDIC Community Banking Study ■ December 2020 2-3 Chart 2.4

New Charters and the Federal Funds Rate, 2009–2019 Number of New Charters Each Year and the Federal Funds Rate as of Each Year End Number Percent 35 2.5 31 Number of New Charters (Le ) 30 Federal Funds Target Rate Lower Limit (Right) 2.0 25

1.5 20

15 13 1.0 11 10 8 0.5 5 5 3 2 1 0 0.0 2009 2010 2011 2012 2013 201420152016201720182019 Sources: FDIC and the Federal Reserve Board (Haver Analytics). Note: Gray bar indicates recession period.

Chart 2.5 New Chartering Rates, 2009–2019 Annual Rates of New Chartering as a Percent of Charters Reporting at Previous Year End Percent 0.5

0.4

0.3

0.2 Average: 2012–2019 +0.1% 0.1

0.0 2009 2010 2011 2012 2013 201420152016201720182019 Source: FDIC. Note: Gray bar indicates recession period.

The New Chartering Rate Remains Low institutions opened in 2012, 2014, or 2016, and during the entire six-year period, only six institutions opened. Late The rate of new charter formation fell to zero in the in the economic expansion new charter formation began aftermath of the financial crisis and Great Recession, to pick up, with 5 new institutions opening in 2017, 8 in and as of 2019 had only barely begun to recover. The last 2018, and 13 in 2019 (Chart 2.4). However, the number of year of substantial new chartering activity was 2008; in new charters in 2019 represented a new chartering rate 2009, the rate of new charter formation set what was at of only 0.2 percent, far below the historical average rate the time a post-1985 record low, and the rate continued of 1.4 percent, which prevailed between 1985 and 2011 to decline until it reached zero in 2012. Almost no new (Chart 2.5). charter formation occurred between 2011 and 2016: no new

2-4 FDIC Community Banking Study ■ December 2020 Table 2.1 Average Annual Rates of Structural Change

Average Annual Rates of Percentage Change in the Number of Charters Between 1985–2011 and Between 2012–2019 Inter- Inter- Intra- Company Other Average Rates of New Net Charter Company Company and Intra- Voluntary Failure Change Because of: Chartering Consolidation Merger Merger Company Closing Merger During 1985–2011 –2 .3 –1 .5 –3 .9 –0 .1 –0 .7 1 .4 –3 .2 the Period: 2012–2019 –3 .3 –0 .6 –3 .9 –0 .2 –0 .2 0 .1 –4 .3 Source: FDIC . Note: Mergers are voluntary . Other Voluntary Closings include institutions that, for example, choose to liquidate without being acquired, or choose to relinquish FDIC insurance . The rates of Net Charter Consolidation, and “Inter-Company and Intra-Company Mergers,” do not equal the sums of their component rates due to rounding .

The Net Consolidation Rate Increased interested in starting banks. It is true that banking became less profitable after the financial crisis, but an important An important fact about consolidation within the banking question is how much of bank profitability post-crisis can industry is that the average annual rate of voluntary be attributed to macroeconomic factors and how much to mergers between 2012 and 2019—combining both mergers other factors, such as regulation. between unaffiliated institutions and those between institutions within the same holding company—was the FDIC research indicates that “[m]ore than 80 percent of the same as the average annual rate of voluntary mergers post-crisis decline in [community bank] profitability can between 1985 and 2011: 3.9 percent (Table 2.1). Moreover, be explained by negative macroeconomic shocks” and that during the period 2012–2019 the average annual rate of the net effects of regulation, business practices, and other failure declined by 0.5 percentage points, while the rate of “structural” factors explain less than 20 percent of the other voluntary closings increased only slightly. Yet the post-crisis decline in profitability.3 average annual rate of net charter consolidation during the period 1985–2011 was 3.2 percent, compared with a rate of It is important to note that while macroeconomic factors 4.3 percent during the period 2012–2019. The increase in appear to explain most of the decline in community-bank net charter consolidation was due to the slow rate of new profitability since the Great Recession and that these factors charters in the latter period. provide a plausible explanation for the low rate of new charter formation, the regulatory environment in which Although a decline in new charter formation following banks operate changed considerably at the same time. For the financial crisis and Great Recession is not entirely detail on how the changed regulatory environment may surprising given the severity of the crisis and recession, have affected community banks, see Chapter 5 of this study. the slow rebound of new charters as the economy recovered is unusual. There are several possible Community Banks Are More Prevalent explanations for it. Macroeconomic factors—such as Than Noncommunity Banks, Although output, interest rates, and unemployment—appear to be Both Groups Continue to Consolidate primarily responsible.2 The possible role of regulatory Among FDIC-insured institutions, community banks are compliance costs in affecting the cost of chartering a new by far the most numerous, and noncommunity banks small bank is discussed in Chapter 5. are the largest by asset size. Also, noncommunity banks have continued to grow their assets at a greater rate than The primary explanation focuses on bank profitability. community banks on average. Both bank types have This explanation maintains that new chartering declined been consolidating since 1986, although community because of the extraordinary decline and weak subsequent banks were less likely to close than noncommunity recovery in bank profitability associated with the financial banks between 2012 and 2019. This section compares crisis and Great Recession. Put simply, this explanation consolidation among community banks with consolidation holds that banking became less profitable after the among noncommunity banks by comparing number of financial crisis and, therefore, fewer investors were institutions, rates of attrition, and average asset growth.

2 Adams and Gramlich; GAO . 3 Fronk .

FDIC Community Banking Study ■ December 2020 2-5 Chart 2.6

Number of FDIC-Insured Institutions, 2009–2019 Number of Charters Noncommunity Banks 9,000 Community Banks 8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0 2009 2010 2011 2012 2013 201420152016201720182019 Source: FDIC.

Chart 2.7 Community and Noncommunity Bank Attrition Rates Between 2012 and 2019 Closings by Type of Closure, as a Percent of Institutions Reporting at Year-End 2011 Percent Community Banks Noncommunity Banks 0 –2 –2 –4 -10 –16 –23 -20 –14 -30 –1 –5 -40 Community Bank Noncommunity Bank -50 Attrition Rate: Attrition Rate: -60 30% 36%

-70 Failed Merged, Within Same Holding Company -80 Merged, Not Within Same Holding Company -90 All Other Voluntarily Closed -100

Source: FDIC. Note: Summation of the percentages by type of closure may not equal the attrition rate due to rounding.

The Number of Community and Noncommunity Although the number of community banks continued to Banks Continues to Decline decline, as of 2019 they were still the most prevalent type of FDIC-insured institution (Chart 2.6). In 2019, 92 percent Between 1985 and 2019 the numbers of both community of all bank charters were held by community banks, and noncommunity banks generally declined, after unchanged from 2011 and up from 87 percent in 1984. increases among both groups between 1984 and 1985. For each group the decline was substantial, especially between Although the number of banks continued to decline, 2012 and 2019 when the number of community banks between 2012 and 2019 community banks were actually dropped by 30 percent and the number of noncommunity less likely to leave the industry than were noncommunity banks by 23 percent. banks. Of the 6,802 institutions that reported as community banks at year-end 2011, just under 30 percent had closed by year-end 2019. In comparison, over the same

2-6 FDIC Community Banking Study ■ December 2020 Box 2.2 Acquirers of Community Banks

Most often, community banks that close do so because they have been acquired by other community banks. Among community banks that ceased operating between 2012 and 2019, just over two-thirds were acquired by other community banks. Even among larger community banks, or those with an asset size between $1 billion and $10 billion, nearly one out of every five that ceased operating was acquired by another community bank (Chart 2.2.1). Chart 2.2.1

Percent of Community Banks at Year-End 2011 That Closed and Were Acquired by Other Community Banks Between 2012 and 2019, by Asset Size Percent 100 89 90 80 70 All Community Banks: 67% 60 65 50 40 32 30 20 18 10 0 <$100 Million $100 Million– $500 Million– $1 Billion– $500 Million $1 Billion $10 Billion Assets of Acquired Community Banks Source: FDIC. Note: Closed community banks failed, voluntarily merged, or voluntarily liquidated.

While most community banks that close do so because they have been acquired by other community banks, more than half of the offices operated by those acquired community banks are acquired by noncommunity banks (Table 2.2.1). This is because banks with larger asset sizes tend to operate more offices compared with smaller banks, and noncommunity banks acquire larger proportions of closed community banks as the asset size of those community banks rises. As shown in Chart 2.2.1, 89 percent of community banks that closed between 2012 and 2019 and had less than $100 million in total assets were acquired by other community banks. However, these relatively small community banks operated two offices each on average, according to data from the FDIC’s Summary of Deposits surveys. Community banks that ceased operating and had between $1 billion and $10 billion in assets, on the other hand, operated 24 offices each on average and were much more likely to be acquired by noncommunity banks.

Table 2.2.1 Offices Acquired and Retained by the Acquirers of Community Banks Between 2012 and 2019 Offices Offices Sold to Number of Offices Initially Retention Rate Offices Closed Type of Acquirer Retained by Other Acquirers Acquired (Percent) by Acquirer Acquirer Institutions Noncommunity Bank 166 5,874 5,086 86 .6 710 78 Community Bank 902 4,727 4,270 90 .3 412 45 Source: FDIC Summary of Deposits data . Notes: The offices initially acquired are those listed as belonging to acquired banks according to their last Summary of Deposits filings . The Summary of Deposits filings of acquiring institutions immediately following mergers are used to determine what happened to the acquired offices . Thus, the Table displays outcomes for acquired offices within the first year or less following an acquisition . These outcomes may be different over a time period longer than one year . period more than 36 percent of the 555 institutions that holding company, as shown in Chart 2.7.4 For details on who reported as noncommunity banks had closed (Chart 2.7). acquires community banks when they merge, see Box 2.2.

In addition to being more likely to close than community 4 banks, noncommunity banks were also more likely to The FDIC defines “community bank” at the holding company level, so separately chartered institutions belonging to the same holding merge with other noncommunity banks within a shared company are either all community banks or all noncommunity banks .

FDIC Community Banking Study ■ December 2020 2-7 Chart 2.8

Average Asset Sizes of Community and Noncommunity Banks, 2009–2019 Average Asset Size ($ Billions) 45

40 Noncommunity Banks $38.4 Billion

35

30

25 82 Times Larger 20

15

10 53 Times Larger 5 Community Banks $0.47 Billion 0 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC.

Average Asset Growth at Noncommunity Banks online and mobile banking, has allowed for the growth of Outpaces Growth at Community Banks noncommunity banks with very large balance sheets. U.S. Gross Domestic Product (GDP) in 2019 was approximately Between 1984 and 2019, noncommunity banks grew 5.3 times larger than GDP had been in 1984. Similarly, substantially compared with community banks, and as the average asset size of community banks in 2019 was of year-end 2019 the average asset size of noncommunity about $470 million, about 5.3 times their average size of banks was 82 times larger than the average asset size of $88 million in 1984. Thus, from 1984 to 2019 community community banks (Chart 2.8). Given the FDIC’s definition banks grew roughly in line with the U.S. economy. The of community bank, however, the growing divergence in average asset size of noncommunity banks in 2019, average size between the two groups should not be entirely however, was more than 38 times their average size in surprising. After all, although the FDIC does not impose 1984, since their growth during that 35-year period far an asset size threshold below which all institutions are outpaced that of the broader economy. The implicit growth considered community banks, the FDIC does impose limits “restriction” on community banks, described above, may on a community bank’s geographic scope, among other be a key factor as to why their share of banking industry things, once the bank reaches a certain asset size, which assets declined slowly after 2011. Between 2012 and 2019, the FDIC adjusts upward over time. As an institution grows the share of banking industry assets held at community its balance sheet, it may grow its geographic footprint. banks declined from 14 percent to 12 percent of the total, Therefore, community banks that grow their balance down from a high of 38 percent in 1984. sheets and expand into new markets may at some point in their growth become noncommunity banks. This Summary implicitly slows down the rate at which the average asset size across all community banks can grow, since fast- The long-term consolidation of the banking industry that growing community banks are more likely to become began in 1986 continued between 2012 and 2019. Bank noncommunity banks. failures contributed less to consolidation as the economy recovered from the financial crisis and Great Recession. On the other hand, noncommunity banks may grow their Mergers made up a greater share of consolidation as assets and footprint very rapidly, raising the average asset failures receded. However, intra-company mergers became size growth rate for all noncommunity banks. The removal less common while inter-company mergers approached of restrictions on both intra- and inter-state branching rates last seen in the 1990s. Because new chartering fell in the 1980s and 1990s, followed by rapid growth in to post-1985 record low rates between 2012 and 2019, the

2-8 FDIC Community Banking Study ■ December 2020 average annual rate of net consolidation increased to Average asset growth at noncommunity banks outpaced 4.3 percent from the rate of 3.2 percent, which prevailed that at community banks between 2012 and 2019. However, between 1985 and 2011. community banks that expand their geographic footprints and their balance sheets may become noncommunity Both community banks and noncommunity banks banks because of their growth, while noncommunity consolidated between 2012 and 2019, although community banks may grow without limit and remain noncommunity banks that existed at year-end 2011 were less likely to banks. Therefore, noncommunity banks are likely to report stop operating between 2012 and 2019 compared with greater rates of average asset growth over time when noncommunity banks. When community banks did cease compared with community banks. operating, more than two-thirds of the time it was because of their acquisition by other community banks.

Box 2.3 Structural Change and the COVID-19 Pandemic

The COVID-19 pandemic could affect the rate of consolidation in important ways. The number of mergers announced publicly fell in early 2020, suggesting that the rate of net consolidation will decline as planned mergers are postponed or canceled. Offsetting this factor, however, is the potential for a rise in bank failures as a result of the pandemic- related economic downturn, particularly if economic recovery is slow. Finally, while the rate of mergers may fall temporarily because of the effects of the pandemic, once the pandemic subsides, mergers could increase, as deals that were postponed are completed.

The rate of net consolidation in the first nine months of 2020 was nearly the same as the rate in the first nine months of 2019. The number of charters declined by 148 during the first nine months of 2019, representing a net consolidation rate of -2.7 percent, and during the first nine months of 2020, the number of charters declined by 144, which equates to a net consolidation rate of -2.8 percent. The number of mergers was 12 fewer during the first nine months of 2020, but there were also five fewer new charters, one more failure, and two more other voluntary closings than there had been in the first nine months of 2019.

More important, the number of merger announcements during the first nine months of 2020 was down 59 percent compared with the number during the first nine months of 2019, suggesting that merger activity would decline later in 2020 and potentially on into 2021. In terms of actual numbers, financial institutions announced 200 mergers during the first nine months of 2019, compared with 82 during the first nine months of 2020, according to data compiled by S&P Global.

Aside from leading to decreases in merger announcements in 2020, the COVID-19 pandemic also led to the termination and postponement of previously announced mergers. In 2019, 11 planned mergers were terminated, compared with 13 terminated mergers in the first nine months of 2020, according to S&P Global data.a In addition, seven planned mergers were postponed or had terms renegotiated and the parties cited the pandemic as one of the factors affecting the decision (Sullivan and Tor).b

a Terminated mergers are not included in the counts of merger announcements . b As of September 30, 2020, five of the seven postponed or renegotiated mergers had been completed .

FDIC Community Banking Study ■ December 2020 2-9

Chapter 3: The Effects of Demographic Changes on Community Banks

The changing demographics of the United States have Counties Can Be Defined by Two Key affected demand for community-bank services, with Demographics: Age and Migration banks seeing changing client bases and therefore changing The term demographic trends refers broadly to major demand for loans as well as other products and services. population characteristics—age, race, sex, marital status, Community banks headquartered in some of the most educational attainment, and many others—and the ways dynamic areas of the United States—those with lower in which they are changing in the nation over time. It is median ages and the highest levels of net migration easy to sense that these trends will affect local economies inflows—are prospering and form an important part of and the community banks that serve them, but it is still the financial community. Community banks in these important to understand how they produce their effects. more dynamic areas experience faster rates of asset and Although there are many different kinds of demographic loan growth, and compared with the community-bank change influencing the U.S. workforce and population, industry as a whole, they are frequently more profitable of particular relevance to community banks are two key and have larger shares of business loans. At the same characteristics: age and migration. Each county in the time, community banks that are serving areas of the United States can be ranked on both its median age and its country with less favorable demographic trends—for net migration rate. example, community banks headquartered in areas with higher median ages and net migration outflows—have Chart 3.1 illustrates these two changes and delimits the fewer opportunities for growth but nonetheless fill a counties of interest in this chapter. The dashed lines split vital role in their local communities. This chapter focuses all counties into quartiles, representing the 25th, 50th, and on the community banks headquartered in the regions 75th percentiles for each age and migration trend. These experiencing the most favorable and the least favorable two sets of quartiles separate counties, and, therefore, the demographic changes, the performance of each group community banks headquartered in them, into 16 groups, relative to the other and to all community banks, and ways but it is only the outermost corners on which this chapter in which the two groups appear to be supporting their local focuses: communities.

• Younger inflow counties are those that are in the In all, the community banks that were headquartered in highest quartile of net migration inflows and the counties where some of the greatest demographic change lowest quartile of median age. was taking place made up 27 percent of all community banks in the United States in 2019—a percentage that has • Older inflow counties are those that are in the highest increased just slightly over time. Put another way, the quartile of net migration inflows and the highest analysis in this chapter encompasses barely more than quartile of median age. a quarter of community banks. It is not meant to ignore the other 73 percent of community banks but, instead, to • Younger outflow counties are those that are in the highlight the differences between groups of community lowest quartile of net migration inflows—which banks facing some of the most extreme demographic in all cases means the community is experiencing situations. Other community banks may be facing population outflows—and the lowest quartile of similar influences on their operations, depending on the median age. demographics of their particular counties, but in any case • Older outflow counties are those in the lowest quartile all community banks can benefit from considering changes of net migration inflows and the highest quartile for in their customer bases. Thus, the analysis as a whole median age. is designed to help all of them better understand their changing customer bases. Although counties not in one of these four groups still are experiencing changing demographic conditions, the best way to illustrate and understand the effect on community

FDIC Community Banking Study ■ December 2020 3-1 Chart 3.1 Median Age and Net Migration Rate, All U.S. Counties Average Annual Net Migration Rate (per 1,000 Residents), 2010–2018 40 Net 30 Younger Inflows Population Older Inflows Inflows 20

10

0 Younger Older -10

-20 Net -30 Younger Outflows Population Older Outflows Outflows -40 25 30 35 40 45 50 55 60 65 Median Age, 2018 Source: 2018 Census American Community Survey. Note: Dashed lines indicate the 25th, 50th, and 75th percentiles of median age and average annual net migration rate.

banks of these two major demographic trends is to focus of life have different credit demands and use different on these four group of counties. banking services.

The United States, like many countries, is growing older as Net migration rate is the other key demographic trend healthcare improves, birth rates decrease, and life spans affecting community banks. People move for many increase. But increases in the average age in the aggregate reasons, among which are school, work, and proximity do not mean that all parts of the country are aging at to family. People also move different distances: within the same rate. Small changes in the national average can the same county, across state lines, and into and out reflect large differences at the county, state, or regional of the United States. Net migration rate is the measure levels. When median age by county, as reported in the 2018 that captures all of this—the number of people moving Census American Community Survey, is delineated into into a county minus the number of people moving out quartiles, counties in the youngest 25 percent are those of it. Although comparing net migration rates can mask where the median age is 36.6 years or below. Counties important differences in why individuals are deciding to in the oldest 25 percent are where the median age is move into or out of a county, net inflows or outflows are 42.5 years or above.1 still an important factor for community banks. Delineating the average annual net migration rate (per 1,000 residents) Map 3.1 displays these oldest and youngest counties, and by county into quartiles shows that “inflow” counties are shows that younger counties are often located more toward those with an average annual migration gain of more than the South and West and also around larger metropolitan 3.7 per 1,000 residents per year, while “outflow” counties areas. Counties with some of the oldest median ages, are those that lose more than 3.7 per 1,000 residents to on the other hand, are frequently located more to the outmigration.2 Northeast, as well as in popular retirement destinations (such as Florida and Arizona) and in more rural areas. Age Map 3.2, which shows the counties with the highest profiles across counties can have important implications inflows and outflows, confirms conventional wisdom for community banks headquartered in those areas and the anecdotes that support it regarding population because people of different ages and in different stages inflows and outflows. Somewhat like counties with the

1 At the state level, the five states with the oldest median age 2 The five states with the highest net migration inflows (descending) (descending) are Maine, New Hampshire, Vermont, West Virginia, and are Florida, Colorado, South Carolina, Arizona, and Washington . Florida . The five states with the youngest median age (ascending) are The states with the highest net migration outflows (descending) are Utah, Alaska, Texas, North Dakota, and Nebraska . Illinois, Alaska, New York, Mississippi, and New Jersey .

3-2 FDIC Community Banking Study ■ December 2020 Map 3.1

Median Age, 2018

Highest Quartile Lowest Quartile Middle Quartiles Source: U.S. Census.

Map 3.2

Average Annual Net Migration Rate per 1,000 Residents, 2010–2018

Highest Quartile Lowest Quartile Middle Quartiles Source: 2018 Census American Community Survey.

youngest median age, counties with the highest net Each county is unique in the factors that affect who lives inflows are larger metropolitan areas or areas popular with there and who moves there, yet between older counties as retirees, like Florida and Arizona. Metropolitan areas, in a group and younger ones as a group there are interesting fact, constitute not only just under 80 percent of inflow and important differences, as there are between inflow counties but also just over 70 percent of younger counties. counties as a group and outflow counties as a group. These Conversely, counties with the highest net outflows are differences affect the community banks headquartered often rural counties. Rural counties constitute almost in the different areas, with some banks experiencing 50 percent of outflow counties and just over 50 percent of an increase in demand and others serving a declining older counties. customer base. Map 3.3 displays counties that exhibit two

FDIC Community Banking Study ■ December 2020 3-3 Map 3.3

Counties in Focus in This Chapter

Younger Inflow Counties Younger Outflow Counties Other Counties Older Inflow Counties Older Outflow Counties

Sources: U.S. Census, 2018 Census American Community Survey. of these key demographic trends simultaneously: oldest furthest corners. Although many community banks are populations with highest outflows, youngest populations clearly serving areas that look similar to banks in the with highest outflows, oldest populations with highest most extreme quadrants with respect to median age and inflows, and youngest populations with highest inflows. average annual net migration rates, between banks in the As noted above, these four kinds of counties are the focus highest and lowest quartiles there are real differences. of this chapter. And the chart strikingly symbolizes one set of differences that Maps 3.1–3.3 depict in a more conventional way: that The Share of Community Banks in Each community banks in metropolitan areas tend to have some County Type Has Been Stable, Tracking of the youngest populations and highest net inflows, while National Consolidation Trends community banks headquartered in rural areas have some of the oldest populations and highest net outflows. The first section of this chapter defined the types of counties where demographic changes are most For the end of each year from 2010 to 2019, Table 3.1 shows pronounced. Though as noted above, all community banks the number and percentage of community banks that were can benefit from considering changes in their customer headquartered in each of the four demographic areas of bases, the rest of this chapter focuses on community banks interest—older inflow counties, older outflow counties, headquartered in highlighted counties shown in Map 3.3. younger inflow counties, and younger outflow counties. Community banks headquartered in each of these four Honing in deeper than Map 3.3 illustrates, Chart 3.2 areas experienced consolidation trends similar to those displays—for each community bank in the country—the for community banking as a whole, and so the number of average annual net migration rate and median age of the charters fell consistently—but the share of community 3 county in which the bank is headquartered. The vertical banks in each of these demographic categories was and horizontal dashed lines in Chart 3.2 represent the roughly stable for the entire eight-year period. All in all, thresholds for the bottom and top quartiles of age and community banks located in these demographic areas net migration rates, respectively. The community banks made up 28 percent of all community banks early in of interest for this chapter are those in the most extreme the decade and 27 percent later in the decade. Shares of quadrants made by the intersecting dashed lines—the community banks in each of the four county types were also remarkably stable during this period. 3 Because statistics are reported at the county level, different community banks in the same county are represented in exactly the same location . Community banks are as of year-end 2019 .

3-4 FDIC Community Banking Study ■ December 2020 Chart 3.2

Scatterplots of All Community Banks, by County Average Annual Net Migration and Median Age Average Annual Net Migration (per 1,000 Residents)

60 Metro Micro Rural 40

20

0

-20

-40 25 30 35 40 45 50 55 60 65 Median Age

Metropolitan Community Banks Micropolitan Community Banks Rural Community Banks Average Annual Net Migration Rate (per 1,000 Residents) Average Annual Net Migration Rate (per 1,000 Residents) Average Annual Net Migration Rate (per 1,000 Residents) 60 60 60 40 40 40 20 20 20 0 0 0 -20 -20 -20 -40 -40 -40 25 30 35 40 45 50 55 60 65 25 30 35 40 45 50 55 60 65 25 30 35 40 45 50 55 60 65 Median Age Median Age Median Age

Sources: FDIC, U.S. Census, 2018 Census American Community Survey. Note: Community banks are as of year-end 2019; Annual Net Migration is from 2010 through 2018; Median Age is 2018. Scale is set on all graphs to include outlier institutions.

Table 3.1 Number and Percentage of Community Banks Headquartered in Key Demographic Areas, Year-End 2010–2019

County Type 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

Younger Number of Institutions 695 665 624 601 563 509 484 467 434 415 Inflows Percent of Community Banks 10 10 10 10 9 9 9 9 9 9

Older Number of Institutions 361 338 320 302 290 273 258 233 217 199 Inflows Percent of Community Banks 5 5 5 5 5 5 5 4 4 4

Younger Number of Institutions 350 341 331 322 308 291 277 267 253 245 Outflows Percent of Community Banks 5 5 5 5 5 5 5 5 5 5

Older Number of Institutions 562 544 532 519 505 494 474 462 445 426 Outflows Percent of Community Banks 8 8 8 8 8 9 9 9 9 9

All Number of Institutions 5,044 4,914 4,737 4,563 4,371 4,169 3,969 3,799 3,631 3,465 Others Percent of Community Banks 72 72 72 72 72 73 73 73 73 73 Total Number of Institutions 7,012 6,802 6,544 6,307 6,037 5,736 5,462 5,228 4,980 4,750 Source: FDIC .

FDIC Community Banking Study ■ December 2020 3-5 Some of the same metrics that were considered in Younger Inflow Counties Chapter 2 of this study (“Structural Change Among The youngest counties with the highest net inflows Community and Noncommunity Banks”) can be considered are arguably some of the most dynamic areas of the in this discussion of community banks headquartered country, and community banks headquartered in these in specific demographic areas. Specifically, net inflow counties are larger than other community banks. In 2019 counties seemed to be a predictor of consolidation activity the median asset size for these community banks was in general. Community banks headquartered in both $313.8 million; the median asset size for community banks younger inflow counties and older inflow counties had a headquartered elsewhere was $206.6 million. Community higher net consolidation rate than did other institutions. banks headquartered in the youngest high-inflow counties And in both types of net inflow county, the most common were also more profitable than other community banks. cause of the decreasing number of individual institutions Throughout the period from 2011 through 2019, the average was outright purchase by another institution, rather community bank in younger inflow counties consistently than failure. It is counterintuitive that consolidation was had a higher NIM than other community banks, by 10 to highest in these counties: they had more customers to 20 basis points. In addition, at these same community serve and were growing faster, and more customers should banks pretax ROA was often higher, usually by 5 to mean higher rates of new bank formation to serve them. 20 basis points. But after mid-2009, the end of the Great Recession, as discussed in Chapter 2, de novo formation was limited. In addition, community banks headquartered in younger

In contrast, community banks headquartered in inflow counties were growing faster than other community older outflow counties experienced lower rates of net banks, as several major parts of banks’ balance sheets consolidation than other institutions. This may be because attested. Between 2010 and 2019, annual asset growth was of the strength of agriculture-focused community banks always faster for the average community bank in younger coming out of the Great Recession.4 Community banks inflow areas than for other community banks. Between in older outflow counties also experienced lower rates of 2012 and 2019 annual deposit growth was greater every outright purchase by another institution. Younger outflow year. And, almost always during the study period, the counties also had lower rates of consolidation than other annual growth rate for loans was higher. institutions earlier in the decade, but by 2015 the rate of consolidation had accelerated some and has been similar to Older Inflow Counties the rest of the United States in recent years. Community banks headquartered in older inflow counties are not as large as their counterparts in younger inflow Community Banks Headquartered in Net counties, but their median asset size of $253.0 million Inflow Areas Had Strong, Profitable Growth made them, too, larger than other community banks Key portions of the balance sheets of community banks located elsewhere. And like their counterparts in younger headquartered in counties with the highest population inflow counties, community banks in older inflow counties inflows indicate that these banks showed strong, profitable experienced stronger growth in key balance sheet metrics growth and continued to support the banking needs than the overall industry. Between 2013 and 2019, the of their local communities. But within inflow areas, annual growth rate for assets at the average community important differences emerge depending on whether the bank in an older inflow county was consistently higher underlying population is older or younger. One can see than the rate for the community-bank industry overall. these differences by focusing on the relationship between Similar trends can be seen in annual loan and deposit demographic trends and the forms taken by asset growth. growth, which have been consistently higher than community banks overall since 2015 and higher more often In the discussion below, the statistics on growth than not during the entire study period. and profitability are calculated using fourth-quarter annualized data for all institutions designated community There is also evidence to suggest that community banks in banks in a given year; assets are not merger-adjusted to older inflow counties had more cash on hand, consistent reflect the ultimate purchaser in preceding years. with anecdotes about retirees keeping amassed assets in FDIC-insured, interest-bearing accounts. The evidence is 4 Chapter 4 has a deeper analysis of agriculture-focused banks .

3-6 FDIC Community Banking Study ■ December 2020 Table 3.2 Commercial and Industrial Loans to Total Assets (Percent)

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 All Community Banks 8 .3 8 .3 8 .4 8 .9 9 .2 9 .3 9 .3 9 .3 9 .8 9 .5 Younger Inflow Counties 10 4. 10 .4 10 .4 11 .0 11 .2 10 .7 10 .5 10 .2 10 .5 10 .4 Older Inflow Counties 6 .1 5 .7 5 .8 5 .5 6 .4 6 .2 6 .2 6 .4 6 .7 7 .2 Source: FDIC .

Table 3.3 Commercial Real Estate Loans to Total Assets (Percent)

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 All Community Banks 28 2. 26 .7 25 .8 27 .0 27 .6 28 .7 30 .1 31 .1 31 .9 31 .2 Younger Inflow Counties 33 4. 31 .2 30 .3 31 .3 31 .8 33 .2 34 .5 35 .2 36 .7 36 .7 Older Inflow Counties 32 6. 30 .6 27 .0 26 .2 26 .5 26 .2 27 .5 27 .9 26 .8 27 .4 Source: FDIC .

Table 3.4 Acquisition, Construction, and Development Loans to Total Assets (Percent)

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 All Community Banks 5 .4 4 .2 3 .8 3 .8 4 .1 4 .4 4 .7 4 .8 5 .0 5 .1 Younger Inflow Counties 7 .7 6 .1 5 .6 5 .8 6 .2 7 .0 7 .0 6 .9 7 .1 7 .1 Older Inflow Counties 7 .7 6 .2 4 .8 4 .6 5 .2 4 .4 5 .0 5 .0 4 .8 5 .1 Source: FDIC . that from 2010 through 2019, the deposit-to-asset ratio For the period 2010 to 2019, Table 3.2 reports the share of for the average community bank headquartered in an C&I loans to assets for community banks headquartered older inflow county was higher than for other community in younger inflow counties, in older inflow counties, and banks. This ratio indicated that these communities might in the community-bank industry as a whole. Community be more deposit-heavy than the average community banks headquartered in younger inflow counties bank elsewhere, which in turn would further support the consistently had a higher share of C&I loans than the anecdotes mentioned above—not only that older customers industry as a whole, but the banks headquartered in older had amassed assets in insured, interest-bearing accounts areas still experiencing net inflows had a lower share of but also that the amount of the savings they had amassed C&I loans than the community-bank industry as a whole, was greater. At the same time, however, unlike the average suggesting possible differences in demand between older community bank in younger areas, the quarterly pretax and younger populations. ROA at the average community bank in older inflow areas was consistently lower than for the average community Table 3.3 reports the share of total assets that CRE loans bank overall. The lower ratio might have been due partly to made up for community banks headquartered in younger the heavy deposit growth and high deposit-to-asset ratio. inflow counties, older inflow counties, and the entire community-bank industry, 2010–2019.5 Community banks Community Banks in Both Younger and Older Net headquartered in younger inflow counties consistently had Inflow Counties Supported Their Communities CRE rates higher than for the community bank industry as Through Business Lending, but Differently a whole. This suggests that community banks in those most dynamic areas were able to support new business growth. Community banks headquartered in net inflow counties, whether older or younger populations, were clearly Table 3.4 reports the share of acquisition, construction, supporting economic growth and the needs of their local and development (C&D) loans to total assets for all communities by issuing business loans. But comparing the community banks and for those headquartered in the shares of certain types of commercial loans makes it clear oldest and the youngest net inflow counties. As with C&I that community banks in younger inflow areas were doing a much larger volume than community banks overall. 5 CRE loans group construction and development loans; multifamily real estate loans; and nonfarm, nonresidential loans .

FDIC Community Banking Study ■ December 2020 3-7 and CRE lending, community banks headquartered in the by 0.5 to 2.5 percentage points, or between only two-thirds youngest net inflow areas consistently had a higher share and 90 percent of the average annual asset growth of other of C&D loans to assets. But whereas in older inflow counties institutions. Starting in 2017, the average community bank the demand for the other two loan categories lagged behind in these areas also saw consistently lower annual loan the demand in other institutions, the C&D loan ratio for growth; and starting in 2014, lower annual deposit growth. older areas was normally at or above the industry average. Yet the slower growth rates and other factors affecting Taken together, these trends suggest that areas with community banks in younger net outflow areas do not population inflows had stronger demand for loan growth appear to have translated into less profitable institutions. and that community banks in those areas were ready to Starting in 2017 the average community bank in a younger serve that demand. Community banks in younger inflow outflow area consistently had a higher quarterly NIM than areas had a higher share of commercial lending than community banks overall. A similar trend is apparent in other institutions. And as noted earlier, areas with older pretax returns. However, both the loans to assets ratio and, populations had more deposits on hand and slower loan starting in late 2014, annual asset growth were lower at the growth—findings that supported anecdotes about the average younger outflow community bank than at other characteristics of an older demographic group. community banks.

Older Outflow Counties Community Banks in Net Outflow Counties Faced Challenges as Demand Growth Faded Many of the issues raised for banks by the demographic The prior section discussed that, between 2012 and 2019, headwind of net population outflows were amplified community banks in net inflow areas grew faster and were in areas with older populations. At year-end 2019, the more profitable than the industry as a whole and some of median asset size at these community banks, at only the ways in which community banks supported commercial $113.8 million, was much smaller than the median asset lending in those areas. In contrast to the higher rate of size at other community banks. And as in outflow areas growth and greater profitability posted by community with younger populations, annual growth rates for assets banks in net inflow areas, growth and profitability were lower for the average community bank in an older among community banks in areas of the country with outflow county than for other community banks—starting net population outflows seem to have been hindered by in 2013, 0.6 to 3.5 percentage points lower. Likewise, from headwinds resulting from this demographic change. 2011 through 2019 the growth rate for loans at the average Even so, differences between outflow areas that serve community bank in an older outflow area was consistently younger populations and outflow areas that serve older lower than for other community banks. The annual growth populations are interesting. As in the previous analysis of rate for deposits displayed the same trend: starting in 2013 inflow counties, the statistics on growth and profitability it was consistently lower at the average community bank in are calculated using fourth-quarter annualized data for all an older outflow area. institutions designated a community bank in a given year; and assets are not merger adjusted to reflect the ultimate The slower balance sheet growth occurring in older purchaser in preceding years. outflow areas seemed to weigh on bank profitability. Starting in 2010, the average community bank in older Younger Outflow Counties outflow areas consistently saw NIMs that averaged 3 to 20 basis points lower than other community banks; lower One way in which net outflows seem to have affected quarterly pretax ROA (though the difference was less stark community banks is by hampering their ability to grow. than for NIMs, and it began in mid-2016); and a higher Between 2014 and 2019, average annual asset growth at the deposit-to-asset ratio (starting in 2010, it was consistently average community bank in younger outflow areas was for higher by roughly 10 to 70 basis points). the most part lower than for other institutions, generally

3-8 FDIC Community Banking Study ■ December 2020 Box 3.1 The Effect of Rural Depopulation on Community-Bank Growth Potential

Even without updated Census designations of rural counties, it is still possible to update the analysis of rural population trends and the implications for banks headquartered in those areas from the 2012 FDIC Study.a Using the 2010 Census county designations for metropolitan, micropolitan, and rural areas but supplementing them with American Community Survey annual population data through 2018, we see that rural depopulation has continued. Between 2010 and 2018, just over 70 percent of rural counties lost population (990 of the 1,353 rural counties had a lower population in 2018 than in 2010). The change from FDIC analyses in 2012 was substantial: in that year, the FDIC reported that 50 percent of rural counties were experiencing depopulation. Furthermore, between the 2012 FDIC study and this study, there was also a further increase in a subset of declining rural counties: rural counties labeled “accelerated declining” because of the quickening pace of their population decline. As of 2019, 300 counties were designated as accelerated rural declining areas, up from 272 in the 2012 study.

In fourth quarter 2019, there were 1,121 community banks headquartered in depopulating rural counties, up slightly from 1,091 at the end of 2011. The 1,121 constituted about 24 percent of all community banks. The reason the number of community banks in depopulating rural counties increased even in the face of continued consolidation in the industry is that more counties began to lose population since 2011. And of the 1,121 community banks headquartered in depopulating rural counties, 391 were headquartered in accelerated declining rural counties.

Concern over the economic effects of depopulation centers on the same issues that previous FDIC analyses highlighted: prime-age workers, those between the ages of 20 and 45, may be moving to seek better opportunities in other places. This can pinch the age distribution of rural counties, and the shrinking tax base that results can increase the fiscal pressure on local governments. In addition, the absence of recent college graduates and other younger workers may make it more challenging for community banks and other local businesses to attract and retain qualified staff, management, and officers, as well as grow their customer bases. The dynamics of out-migration and depopulation risk becoming self-reinforcing, a risk highlighted in the prior FDIC studies.

The median asset size of a community bank in rural declining areas has been much smaller than the median asset size of a community bank headquartered in other areas. The 2012 FDIC Study found that from 2001 to 2007 community banks located in rural depopulating counties reported lower pretax returns than did community banks in other areas—but the study also found that from 2007 to 2011 these community banks had higher earnings. During the latter period, the performance success of depopulating rural banks relative to other institutions was mostly attributable to rural banks’ dependence on agriculture, a sector that remained particularly strong throughout the Great Recession. The Great Recession largely hit metro areas, whereas the agriculture industry was spared major economic shocks. During the study period banks in rural declining areas consistently had a much higher share of agricultural loans to total assets, ranging from 14 to 19 percent of total assets and always at least triple the share of community banks headquartered in other areas. Agriculture-focused rural banks performed better during, and recovered more quickly from, that recession.

The period between 2011 and 2019 saw rural banks in depopulating areas continue to report higher earnings, and quarterly NIM was persistently around 5 basis points higher at these banks than at other institutions. This is once again attributable to the focus on agriculture lending at many of these institutions. Some of this advantage, however, eroded over time because of the fall in global commodity prices that began in 2014. Thus, although pretax returns recovered from the Great Recession more quickly at rural community banks than at other institutions and were higher initially, the situation reversed in 2015. Even so, going into 2019, community banks that specialized in agriculture were more profitable than community banks that were simply headquartered in rural communities. (See Chapter 4 of this study for details on agricultural specialists.) continued on page 3-10

a For an analysis, see FDIC Community Banking Study (2012), Chapter 3 . Anderlik and Cofer (2014) also addresses the issue of rural depopulation .

FDIC Community Banking Study ■ December 2020 3-9 Box 3.1, continued from page 3-9

From early 2014 through 2019, the demographic headwinds of rural depopulation weighed more heavily on other parts of community-bank balance sheets. Asset growth was weaker at community banks in rural declining regions than at other community banks: annual growth rates for assets were consistently between 1.5 and 3 percentage points lower than they were for other banks. During the same period, the average community bank in rural declining areas saw slower loan growth and slower deposit growth than the average community bank. Starting in late 2013, growth rates for both loans and deposits generally ran 1 to 3.5 percentage points lower, or roughly a half to two-thirds of the growth experienced by other institutions.

In summary, these trends indicate a continuation of findings from the 2012 Study. The performance reported here of depopulating rural banks relative to other community banks is somewhat surprising because the agricultural sector, which many of these banks service, faced low commodity prices during the latter part of the period between 2012 and 2019. Until the appearance of COVID-19 (discussed more fully in Box 3.2), the outlook for rural depopulation was for demographic conditions to continue their long-term trend of deterioration, with more migration out of rural counties, more pinching of the distribution of ages (with prime-age workers leaving), and some of the fastest-growing rural counties set to be upgraded to micropolitan areas in the 2020 Census.

Community Banks in Net Outflow Areas Do Not for all community banks. This suggested that in the Have Similar Commercial Lending Portfolios to coming years, perhaps the commercial loans demanded by Other Community Banks a younger population would help support economic growth in their areas. Partly because of the demographic headwinds outlined above, community banks headquartered in net outflow Table 3.6 displays the share of CRE loans for all community counties often had lower commercial lending volumes banks and for the institutions headquartered in older than other institutions. Table 3.5 reports the share of outflow and younger outflow counties. Between 2010 and C&I loans to total assets for all community banks and 2019 community banks headquartered in younger outflow for the institutions headquartered in older outflow and counties had CRE to asset ratios that were near—but younger outflow areas during the period 2010 through always below—the ratios of the industry as a whole. 2019. Community banks headquartered in older outflow Institutions in older counties, however, had CRE volumes counties consistently had a lower share of C&I loans than much lower than those of the industry, suggesting less other institutions. Community banks in younger outflow underlying demand for these types of commercial loans, areas showed a slightly different trend. Although they which in turn may have been an additional headwind had a lower C&I loan share in the years immediately after pushing against continued community-bank growth in the Great Recession, starting in 2012 their share steadily those locations. climbed and, starting in 2017, was higher than the share

Table 3.5 Commercial and Industrial Loans to Total Assets (Percent)

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 All Community Banks 8 .3 8 .3 8 .4 8 .9 9 .2 9 .3 9 .3 9 .3 9 .8 9 .5 Younger Outflow Counties 8 .0 7 .8 8 .1 8 .3 8 .9 8 .8 9 .1 9 .5 10 .3 10 .6 Older Outflow Counties 8 .1 7 .8 7 .6 7 .9 7 .9 8 .0 8 .1 8 .3 8 .5 8 .5 Source: FDIC .

Table 3.6 Commercial Real Estate Loans to Total Assets (Percent)

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 All Community Banks 28 2. 26 .7 25 .8 27 .0 27 .6 28 .7 30 .1 31 .1 31 .9 31 .2 Younger Outflow Counties 28 0. 25 .9 25 .7 25 .8 26 .6 28 .0 30 .0 30 .4 30 .6 30 .1 Older Outflow Counties 17 4. 16 .3 15 .5 15 .8 16 .4 17 .1 18 .0 19 .0 19 .9 19 .8 Source: FDIC .

3-10 FDIC Community Banking Study ■ December 2020 Table 3.7 Acquisition, Construction, and Development Loans to Total Assets (Percent)

2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 All Community Banks 5 .4 4 .2 3 .8 3 .8 4 .1 4 .4 4 .7 4 .8 5 .0 5 .1 Younger Outflow Counties 5 .6 4 .6 4 .2 4 .1 4 .3 4 .7 5 .0 5 .0 4 .9 4 .8 Older Outflow Counties 2 .6 2 .1 1 .9 1 .9 2 .1 2 .4 2 .6 2 .9 3 .2 3 .1 Source: FDIC .

Table 3.7 reports the share of C&D loans to total assets groups of banks experienced lower deposit growth rates for all community banks and for the institutions than other parts of the industry, the deposit to asset headquartered in older outflow and younger outflow share of community banks in older outflow areas was counties. From 2010 through 2017, community banks in significantly higher than for other community banks, younger outflow counties had C&D loan ratios above those suggesting that retirees were continuing to keep their of the industry as a whole, suggesting that these banks money with local banks. were able to support the economic expansion. In 2018 and 2019, however, the levels in these counties slipped below Summary those of the industry. Levels of C&D loans in community banks headquartered in older outflow counties was less Community banks serve customers in their local encouraging: the share of C&D loans at these institutions geographic areas, and long-term population trends was much lower than for the industry as a whole—in some affect the individuals located in an area and the services years, almost half as low—though the level of such loans those customers demand. In areas of the country that are has risen steadily since 2012. arguably most thriving—younger with net population inflows—community banks are growing quickly and

This group of trends as a whole suggests that community profitably and are supporting communities with C&I and banks headquartered in areas experiencing population CRE loans to help areas continue to grow. There is some outflows were less profitable and slower growing than concern, however, whether some of the areas experiencing other community banks. Worth noting, though, is the net outflows will be able to continue to grow; banks in difference in deposit growth rates between community those areas have slower growth and lower commercial banks headquartered in older outflow areas and those lending portfolios, both of which could weigh down headquartered in younger outflow areas. Although both community banks in those areas and possibly feed into higher consolidation rates in the future.

Box 3.2 Net Migration Rates and the COVID-19 Pandemic

The emergence of the COVID-19 pandemic was an unexpected shock that affected the economy with immense speed and force. Unlike other areas of the economy that the pandemic has disrupted, however, demographic trends are slow to change: because the U.S. population is so large, demographic trends in this country normally take decades to develop and make their economic mark. Thus, changes in the population data that are due to the pandemic are not likely to be seen for some time. Even if in retrospect there is a clean break in some demographic trends beginning in 2020, most likely the changes will not appear in population data for a number of years.

One issue worth monitoring for its potential effect on demographics over the longer term is remote working. The government-mandated requirement for social distancing to reduce infection has led to a temporary increase in telework in many industries. If this increase in telework becomes a permanent feature for segments of the workforce, it may allow workers to move to locations outside major cities and still be productive. They may choose to relocate to areas with more open space or a lower cost of living, which could increase migration overall as well as changing which counties are the areas of highest inflows. Additional telework flexibility could even reverse long-standing trends of inflows into the largest urban areas.

FDIC Community Banking Study ■ December 2020 3-11

Chapter 4: Notable Lending Strengths of Community Banks

Community banks provide their local communities CRE Lending with valuable products and services, including offering Throughout the United States CRE lending is an important various loan products to business owners and developers, function performed by banks of all sizes, including small businesses, and farms. As discussed in Chapter 3, community banks. As of year-end 2019, banks held community banks are successful in areas that are $2.3 trillion in CRE loans, an amount that gave them a experiencing a population inflow—areas filled with local significant presence in the broader financial industry. small businesses. But community banks also continue to Community banks in aggregate held almost one-third of meet the credit needs of less economically vibrant areas, this amount—$690 billion—despite having only a small such as rural counties experiencing population outflows. share (12 percent) of the banking industry’s total assets. Community banks tend to focus on loans as relationships, Moreover, as Chart 4.1 indicates, community banks’ originating loans that require local knowledge, a greater share of CRE loans has been relatively stable since 1989 personal touch, individual analysis, and continued even while their share of total banking industry assets administration rather than loans that can be made was declining. according to a formula.

Community banks’ participation in CRE lending is In this chapter, we discuss three lending areas that are widespread. Almost all 4,750 community banks hold at particularly important for community banks: CRE lending, least some CRE loans, and many have substantial CRE small business lending, and agricultural lending. Though loan portfolios. More than one-fifth of community by definition community banks tend to be relatively small, banks have CRE loan portfolios equal to or greater than in each of these areas their lending far exceeds their three times their amount of capital—above the share of aggregate lending share: community banks represent community banks that have substantial portfolios in any 15 percent of the industry’s total loans but 30 percent of other loan type. its CRE loans, 36 percent of small business loans, and 70 percent of agricultural loans.

Chart 4.1

Community Banks’ Share of the Banking Industry’s Assets and CRE Loans, 1989 to 2019 Percent 40 Assets CRE Loans

35

30

25

20

15

10

5

0 1989 1991 1993 1995 1997 1999 2001 2003 20052007200920112013201520172019

Source: FDIC.

FDIC Community Banking Study ■ December 2020 4-1 Chart 4.2 Chart 4.3 Commercial Real Estate Loans Held by FDIC-Insured Commercial Real Estate Loans Held by FDIC-Insured Community Banks, Year-End 2011 Community Banks, Year-End 2019 (Total $521 Billion) (Total $690 Billion) Residential C&D, $20 Billion, Residential C&D, $36 Billion, 4 percent of total 5 percent of total Nonresidential C&D, $63 Billion, Nonresidential C&D, $77 Billion, 12 percent of total 11 percent of total Nonfarm/Nonresidential - Nonfarm/Nonresidential - Owner Occupied, Owner Occupied, $183 Billion, $196 Billion, 35 percent of total Multifamily, $58 Billion, 29 percent of total Multifamily, $102 Billion, 11 percent of total 15 percent of total

Nonfarm/Nonresidential - Nonfarm/Nonresidential - Non Owner Occupied, Non Owner Occupied, $198 Billion, 38 percent of total $279 Billion, 40 percent of total Source: FDIC. Source: FDIC. Note: C&D stands for acquisition, construction, and development loans. Note: C&D stands for acquisition, construction, and development loans.

Community banks provide various types of CRE financing. financing of investor-owned properties as opposed to Charts 4.2 and 4.3 show the types of loans that constituted owner-occupied properties, that is, properties whose total CRE loans held by community banks at the time of owners use the property to operate a business. Loans the 2012 FDIC Community Banking Study (year-end 2011) to finance construction of properties increased only and as of year-end 2019. The three main components of modestly in dollar amount between 2011 and 2019; the CRE loans are loans secured by nonfarm, nonresidential modest increase likely reflects moves away from this type properties; loans for the acquisition, construction, and of lending in the wake of the construction-loan stress development of real estate (C&D); and loans secured experienced by many banks during the Great Recession.2 by multifamily properties. Loans secured by nonfarm, Between 2011 and 2019, construction loans’ share of nonresidential properties are further divided into two community banks’ total CRE loans remained steady at groups according to whether the property is occupied by 16 percent. an owner or by a non-owner. C&D loans are further divided into two groups according to whether they are secured Community Banks Are Active Lenders Across the by nonresidential construction projects or by 1–4 family Spectrum of CRE Industries and Are Key Lenders in residential projects. Small Communities

Banks’ Call Reports categorize CRE loans by segment, Between 2011 and 2019, the balance of CRE loans held such as the three just discussed: multifamily property by community banks increased from $521 billion to loans; C&D loans; and nonfarm, nonresidential loans. $690 billion. Although all types of CRE loans grew in dollar This categorization provides some insight into the type of amounts, the mix shifted toward multifamily loans, that property that secures a CRE loan, but for banks’ portfolios is, loans secured by rental properties.1 of nonfarm, nonresidential loans—that is, CRE loans that are not multifamily or C&D loans—the Call Report does Multifamily loans represented 11 percent of community not indicate the type of business or industry that uses the banks’ CRE loans in 2011 and rose to 15 percent in 2019. existing commercial property. This shift reflects growth in multifamily lending in the broader financial industry during a period when Other CRE industry data, however, suggest that regional multifamily living became increasingly popular. and local banks, many of which are similar in profile Nonfarm, nonresidential loans represented 73 percent to community banks, are active lenders to multiple of CRE loans in 2011 and dropped to 69 percent in 2019. industries. According to real estate firm Real Capital And as the chart shows, nonfarm, nonresidential loans Analytics, regional and local banks lend across the shifted toward those secured by non-owner-occupied properties. This shift suggests an increased focus on the 2 Banks that held high levels of C&D loans before the Great Recession 1 Multifamily loans are those secured by properties with five or more failed at a higher rate than those that did not (2012 FDIC Community housing units . Banking Study pp . 5–15) .

4-2 FDIC Community Banking Study ■ December 2020 Chart 4.4 CRE Loan Origination Market Share by Lender Type and Industry, 2012 to 2019

Regional/ National International CMBS Financial Government Insurance Private/ Local Bank Bank Bank Agency Other

Apartment 12% 10% 4% 6% 5% 55% 7% 1% 1% Hotel 17% 13% 10% 39% 14% 5% 1%

Industrial 25% 22% 7% 19% 7% 1% 17% 2%

O ice 14% 20% 12% 25% 12% 1% 15% 1%

Retail 21% 14% 8% 33% 7% 1% 14% 1%

Senior Housing 16% 16% 5% 5% 13% 42% 1%

Source: Real Capital Analytics. Note: Financial includes financing companies, mortgage real estate investment trusts, and institutional debt funds.

Chart 4.5

Shares of CRE Loan Originations in Major Metro, Secondary, and Tertiary Markets by Lender Type, 2019

Regional/ National International CMBS Financial Government Insurance Private/ Local Bank Bank Bank Agency Other

Major Metros 17% 19% 13% 16% 12% 12% 10% 1%

Secondary 16% 17% 6% 17% 9% 23% 11% 1%

Tertiary 28% 15% 3% 17% 8% 22% 6% 1%

Source: Real Capital Analytics. Note: Real Capital Analytics bases its market size categorizations on the amount of lending in a given market, not on population. Financial includes financing companies, mortgage real estate investment trusts, and institutional debt funds.

spectrum of industries that operate CRE. Chart 4.4 In addition to lending across industry types, community shows the distribution of CRE loan originations from banks have been active CRE lenders across all sizes of 2012 through 2019 according to use of the underlying markets, and are particularly prominent in smaller property. Regional and local banks’ market share has communities. According to Call Report data, community been significant in several property types, including banks headquartered in rural areas and small metropolitan industrial and retail. During the period covered, these areas in 2019 held 67 percent of CRE loans held by all banks originated 25 percent and 21 percent of the banks headquartered in those smaller geographic areas. In dollar volume of industrial and retail loans originated, larger metropolitan areas, the share of CRE loans held by respectively—notable market shares, given the range of community banks is lower, but still material: 28 percent of industry lenders. total CRE loans of all banks headquartered in these areas.

FDIC Community Banking Study ■ December 2020 4-3 In contrast, community banks’ share of non-CRE loans is of community banks have entered multifamily lending for only 9 percent.3 the first time. Of the 4,750 community banks in 2019, 474, or 10 percent, had multifamily loans on their books in 2019 Although Call Report data are based on the location of but had none at year-end 2011. In comparison, during the a bank’s headquarters rather than the location of the same period very few community banks newly entered property securing the loan, other CRE industry data other lending businesses. For example, only 59, slightly are based on property location and they, too, suggest more than 1 percent of community banks, newly entered that banks similar in profile to community banks are C&I lending between year-end 2011 and year-end 2019. significant sources of CRE financing in smaller markets. As Chart 4.5 shows, according to Real Capital Analytics, The increase in multifamily lending pushed community regional and local banks provided 28 percent of CRE banks’ average ratio of multifamily loans to capital from financing in smaller markets in 2019, a material market 27 percent at year-end 2011 to 39 percent at year-end share compared with the shares of other lenders.4 2019. The average ratio of multifamily loans to capital increased in almost all states between 2011 and 2019. But Community Banks Became More Involved in some states, multifamily lending is more important in Multifamily Property Lending After the to community banks than in others. In several states Previous Study in the northeast, such as New York, New Jersey, and Massachusetts, and in California, community banks’ By year-end 2019, the volume of multifamily mortgage average ratios of multifamily loans to capital at year-end loans had almost doubled from its level in 2011.5 At 2019 were well above the national average. The higher year-end 2019 multifamily mortgage loans in the United ratios are consistent with the above-average prevalence States totaled $1.6 trillion. These loans are held by of multifamily living in these states.7 various intermediaries such as banks and life insurance companies, and are also held in agency commercial- Community Banks That Specialize in CRE Lending mortgage-backed securities. Significant growth in Became More Prominent in the Years After the multifamily mortgage loans reflects the increase in Previous Study multifamily housing stock, and the increase in preference for renting following the Great Recession and its associated Community banks of all lending specialties provide housing crisis. Nationally, from 2011 to 2019 the number of CRE financing; however, the share of community banks renter households grew more than 13 percent, while owner considered to be CRE specialists has grown.8 The 2012 households increased only 6 percent.6 Community Banking Study found that at the highest point of their share of all community banks, in 2007, CRE As the volume of multifamily loans industry-wide grew, specialists had come to constitute almost 30 percent of the share held by banks kept pace. These institutions held community banks. The share declined from 2008 to 2012, approximately one-third of the $1.6 trillion in multifamily amid the economic slowdown and CRE market stress in mortgages outstanding at year-end 2019, up slightly the few years following the Great Recession, but after that from 2011. As of year-end 2019 community banks in the share recovered somewhat and then stabilized. As aggregate held a small share—22 percent—of all banks’ of year-end 2019, CRE lending specialists accounted for multifamily loans, but since the prior study a large number 26 percent of all community banks (Chart 4.6).

3 Market size is determined according to data from the U .S . Census . Notably, while CRE specialists accounted for, on average, “Larger metropolitan areas” are those designated by the U .S . Census as metropolitan statistical areas—those that have at least one about one-quarter of community banks from 2011 urbanized area of 50,000 or more inhabitants . “Small metropolitan to 2019, their share of community banks’ assets and areas” are those designated by the U .S . Census as micropolitan statistical areas—those that have at least one urban cluster of at least 10,000 but less than 50,000 population . “Rural areas” are those not in 7 2018 American Community Survey, 1-Year Estimates, U .S . Census a metropolitan or micropolitan statistical area . Bureau . The four states mentioned in the text have a higher 4 Real Capital Analytics bases its market size categorizations on the percentage of total housing identified as containing five or more amount of lending in a given market, not on population . housing units than the national percentage . 5 Federal Reserve, Report Z 1. – Financial Accounts of the United 8 As shown in Appendix A, CRE specialists hold construction and States, March 2020 . development (C&D) loans greater than 10 percent of assets OR total 6 U .S . Census Bureau, Current Population Survey/Housing Vacancy CRE loans (C&D; multifamily; and nonfarm, nonresidential loans) Survey, March 10, 2020 . greater than 30 percent of total assets .

4-4 FDIC Community Banking Study ■ December 2020 Chart 4.6

CRE Lending Specialists’ Share of the Number of Community Banks, Community Banks’ Assets, and Community Banks’ CRE Loans, 1989 to 2019 Percent 70

60 Banks Assets 58 50 CRE Loans 40 41 30 26 20

10

0 1989 1991 1993 1995 1997 1999 2001 2003 20052007200920112013201520172019

Source: FDIC. Note: Data as of year end.

Table 4.1 CRE Lending Specialists’ Share of the Number of Community Banks, Community Banks’ Assets, and Community Banks’ CRE Loans, by Market Size, 2011 and 2019

Rural/Micro Metro Total 2011 2019 2011 2019 2011 2019 Percent of Community Banks 10 13 36 38 24 26 Percent of Assets 18 26 40 47 33 41 Percent of CRE Loans 33 41 57 63 51 58 Sources: FDIC, United States Census Bureau . Note: Data as of fourth quarter . Market size is determined according to data from the United States Census Bureau . A metropolitan statistical area must have at least one urbanized area of 50,000 or more inhabitants, while each micropolitan statistical area must have at least one urban cluster of at least 10,000 but less than 50,000 inhabitants . Rural areas are those not in a metropolitan or micropolitan statistical area .

CRE loans increased to an outsized degree. As of year-end The CRE Credit Environment Was Favorable 2019, CRE specialists accounted for 41 percent of aggregate at the Start of 2020 community-bank assets and 58 percent of aggregate For much of the period since the prior study, community community-bank CRE loans. banks, like much of the CRE finance industry, experienced a benign credit environment. Delinquency rates among Community-bank CRE specialists have maintained their community banks’ CRE loan portfolios declined for nine significance across different sizes of geographic markets. consecutive years, from 2010 through 2018, before flattening Not surprisingly, these specialists are prominent in at a low level in 2019. As of first quarter 2020, the average larger geographic markets, that is, where population CRE loan delinquency rate was about 1 percent, much densities and high volumes of real estate provide lower than the peak of more than 7 percent reached in first lending opportunities for all types of lenders. However, quarter 2010. community-bank CRE specialists are also important providers of CRE financing in small communities. Despite As important providers of CRE financing, community accounting for only 13 percent of the number of community banks will be among those lenders interested in CRE banks headquartered in rural/micro markets, community- market dynamics in the years ahead. As 2020 began, the bank CRE specialists held 41 percent of community banks’ long economic expansion had been a positive backdrop CRE loans in these markets in 2019, up from 33 percent in to conditions in the CRE market. However, the landscape 2011 (Table 4.1).

FDIC Community Banking Study ■ December 2020 4-5 weakened significantly with the COVID-19 pandemic, and lending sources for business owners, property developers, economic stress likely will be a headwind holding back and investors. Community banks hold a larger amount the performance of numerous CRE property types. In of CRE compared with their overall industry asset share. addition, potential shifts in preferences for certain types They fund a wide variety of properties in locations of real estate over others may change the CRE lending throughout the country, and as the demands of their environment. For insights into CRE market conditions and communities change, their CRE lending changes to meet the COVID-19 pandemic, see Box 4.1. the need. Moreover, while some community banks may be considered CRE specialists because of the share of CRE CRE Lending: Summary loans in their portfolios, most community banks hold some CRE loans, supporting the premise that whatever a Despite challenges in CRE markets or the economy, community bank’s business strategy, the bank is focused community banks have been and continue to be CRE on the various needs of its community.

Box 4.1 CRE and the COVID-19 Pandemic

The COVID-19 pandemic has substantially altered the landscape of CRE markets in the United States. As businesses and industries reevaluate their use of space, questions have emerged about how CRE will be used, the amount needed, and the geographic implications.

Nationally, rents have declined and vacancy rates have increased in most property types since the onset of the pandemic, and projections call for continued weakness. For example, by the end of 2021, real estate firm CoStar projects that vacancy rates will increase by 20 percent or more in most property types.a As the pressure on rents and occupancy rates continues, ultimately CRE property prices are expected to show the strain. According to CoStar, prices in most property types are expected to decline by double digits into 2021 and to recover slowly from the COVID-19 pandemic (see Chart 4.1.1).

The effects of the COVID-19 pandemic have affected property types in different ways. As the pandemic emerged with government-mandated business and travel restrictions, immediate stress was felt in lodging and retail, as hotels, restaurants, and stores closed. Foot traffic at discretionary retail stores fell to near zero. The national hotel occupancy rate dropped to a low of 21 percent in April 2020, from a pre-pandemic monthly average in 2019 of approximately 66 percent.

Chart 4.1.1

Projected Percent Change in Property Prices by Quarter and Property Type, First Quarter 2020 Through Second Quarter 2022 Forecast Change in Prices Indexed to the Pre-Pandemic Level (First Quarter 2020) Percent 5

0

-5 Industrial Retail -10 Multifamily O ice

-15 1Q20 2Q20 3Q20 4Q20 1Q212Q213Q214Q211Q222Q22 Source: CoStar. Note: Data are quarterly figures forecasted as of second quarter 2020.

continued on page 4-7

a CoStar forecast as of second quarter 2020 .

4-6 FDIC Community Banking Study ■ December 2020 Box 4.1, continued from page 4-6

Companies’ use of office space slowed, and many cities’ office markets may experience challenges in long-term demand as companies reevaluate their space needs. Office markets in some geographies may be strained more than others, depending on various factors such as long-term adoption of telework, challenges in cities highly dependent on public transportation, and the path of COVID-19 as a long-term health crisis.

Depending on the depth of economic contraction, the pace of recovery, and living preferences among renters, multifamily markets also may face headwinds. Some multifamily properties may be strained by delays in the payment of rents and by the recent large increase in multifamily supply in some markets.

Overall, CRE market weakness may manifest itself in the credit quality of CRE loan portfolios. Credit quality may suffer as economic strain from the COVID-19 pandemic increases vacancy rates, reduces properties’ cash flows, and— potentially—hinders loan repayment ability.

Small Business Lending (22 percent) and credit unions (6 percent).11 Although noncommunity banks may provide a larger portion of Small businesses play a key role in the economy, making up small business loans by dollar amount, figures for overall the vast majority of all businesses by count and employing market share and the small-business-loans portion of approximately 48 percent of the private sector workforce.9 total business loans make it clear that community banks In addition, they are an important part of their community, tend to lend primarily to small businesses. An analysis of not only by providing services and products but also by Call Report data in conjunction with responses to the 2018 supporting local causes and charities. And just like large FDIC Small Business Lending Survey and loan origination corporations, they need to borrow funds for a number of data from the Small Business Administration (SBA) shows reasons, including to add to working capital and inventory, that community banks are key providers of loans to small to finance accounts receivable, and to purchase properties local businesses and are key resources for small businesses that house their businesses. needing credit.

In contrast to large corporations, which may be more Call Report Data Are Helpful but Do Not likely to turn to the capital markets, small businesses Show the Full Story more frequently turn to banks for credit, particularly if the business owner has a relationship with a lender.10 Call Reports are the primary source for analyzing growth Banks provide approximately 44 percent of small business and changes in banks’ small business loans. According financing, considerably higher than online lenders to Call Report data, small business loans grew from $599 billion at year-end 2011 to $645 billion at year- end 2019, for an average annual rate of loan growth of Study Definitions 0.98 percent. This growth rate is considerably less than the average annual business loan growth rate of 6.8 percent In this study, business loans are all C&I loans and for the banking industry. Growth in small business loans all nonfarm, nonresidential loans. Business loan was solely in small C&I loans, since small nonfarm, growth reflects growth in all C&I and nonfarm nonresidential loans fell from $316 billion to $275 billion nonresidential loans for all banks. during the period in question (Chart 4.7). Yet despite the The Call Report defines small business loans as slow growth trends, community banks’ share of small all C&I loans less than $1 million and nonfarm, business loans as of year-end 2019 continues to be larger nonresidential loans less than $1 million. This dollar than their overall share of the banking industry’s total limit was established in 1993 when this category loans. Community banks hold 36 percent of total small was added to the Call Report. This study uses this definition. business loans, which is double their share of the banking industry’s total loans (15 percent).

9 Federal Reserve Banks . 10 Ibid . 11 Ibid .

FDIC Community Banking Study ■ December 2020 4-7 Chart 4.7 Small Business Loans by Type, 2011–2019 $ Billions 700 Total Small C&I Total Small Nonfarm Nonresidential

600

500

400

300

200

100

0 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC. Note: Small C&I and small nonfarm nonresidential loans include all loans with origination amounts less than $1 million.

Community banks have, however, seen their share of do not have to build a relationship or take additional small business loans decline since 2011; it shrank from measures to learn about the business owner or the business 42 percent at year-end 2011 to the aforementioned itself. Using this model, noncommunity banks originate 36 percent at year-end 2019 (Chart 4.8). This decline may and hold more loans under $100 thousand than loans be due to two factors: business consolidation and the between $100 thousand and $1 million. Community banks, typical size of loans made by noncommunity banks in on the other hand, hold a greater share of loans between contrast to community banks. The first factor (business $250 thousand and $1 million than loans under $250 consolidation) would affect community banks’ lending if thousand (Chart 4.9). Community banks, therefore, focus a decline in the number of small businesses (using small on larger loans that require greater “touch” or interaction business employment as a proxy) meant a reduced demand and analysis—loans that build a relationship between bank for small business loans.12 The period when the decline in and borrower. the share of small business loans occurred was the period when small business employment declined (dropping The fact that community banks originate larger small from 50 percent in 2011 to 47 percent in 2017).13 During business loans than noncommunity banks leads us to that period, community banks’ share of small C&I loans an additional hypothesis as to the reason for the decline declined from 32 percent to 25 percent—but their share of in community banks’ share of small business lending. small nonfarm, nonresidential loans remained relatively These larger loans would include those that exceed stable at 51 percent. $1 million—the maximum small business loan limit (see the sidebar “Study Definitions” for more details). The The second possible explanation for the decline in reason for the limit was that there was no one measure to community banks’ share of small business loans is the use in identifying a small business: Is the determination size of loans originated. Size of loan differs significantly based on revenue? On number of employees? On capital between community and noncommunity banks. invested by the business owner? Setting a loan limit for Noncommunity banks tend to help small businesses by reporting purposes gave bankers a simple way to identify offering business credit cards rather than other types of small business loans and ensure uniformity in Call Report working capital or CRE loans. To determine whether to filings. To provide support for our hypothesis that the extend a loan, noncommunity banks use scoring models or decline in community banks’ share of small business other tools, and by using such technology, bank personnel lending between 2011 and 2019 may be partly due to the size of some of their larger loans—with loans exceeding 12 Brennecke, Jacewitz, and Pogach . $1 million not being categorized as small business— 13 Ibid .

4-8 FDIC Community Banking Study ■ December 2020 Chart 4.8

Community and Noncommunity Banks’ Share of Small Business Loans, 2011–2019 Percent Noncommunity Bank Community Bank 100 90 36% 80 42% 70 60 50 40 64% 30 58% 20 10 0 2011 2012 2013 2014 20152016201720182019 Source: FDIC. Note: Small loans to businesses include commercial and industrial loans less than $1 million and nonfarm, nonresidential mortgages less than $1 million.

Chart 4.9 Community Bank and Noncommunity Bank Small Business Loans by Dollar Size, Year-Ends 2011 and 2019 $ Billions 160 2011 2019 140 120 100 80 60 40 20 0 < $100 $100–$250 $250 < $100 $100–$250 $250 Thousand Thousand Thousand- Thousand Thousand Thousand- $1 Million $1 Million Community Banks Noncommunity Banks Source: FDIC.

we look at responses to the FDIC’s 2018 Small Business growing. Small C&I loans as of year-end 2019 represent Lending Survey and to loan origination data from the SBA’s 43 percent of total C&I loans at community banks, whereas 7(a) loan program. small C&I loans currently represent only 14 percent of total C&I loans at noncommunity banks. What Do Bankers Consider to Be Small Business Lending? In 2018, the FDIC, with assistance from the U.S. Census Bureau, conducted a small business lending survey Although community banks’ share of total business loans is (referred to hereafter as “Lending Survey”) with direct declining, within total business loans at community banks responses from banks. Several questions centered on the share represented by small business loans has been the topic of what the banker considers a small business

FDIC Community Banking Study ■ December 2020 4-9 Chart 4.10

Adjusted Small Commercial and Industrial Loans, 2012–2019 $ Billions 100 90 80 70 60 50

40 Adjusted Small C&I Based on Lending Survey 30 Reported in the Call Report Small C&I 20 10 0 2012 2013 2014 20152016201720182019 Source: FDIC. Note: Represents commercial and industrial loans at banks with total assets less than $1 billion. loan. The responses indicate that many bankers do not SBA Loan Originations Also Support the Belief That define a “small business loan” as a loan to a business Community Banks Focus on Small Business Lending with an origination amount less than $1 million, as Community banks are also key players in the Call Reports define it. Rather, bankers consider the SBA-guaranteed 7(a) loan program, which guarantees “ownership structure, number of employees, business loans originated up to $5 million.15 Between 2011 and focus, and ownership characteristics” of the borrower 2019, community banks saw their share of SBA 7(a) loan to determine whether a loan is to a small business; often originations increase from $5.7 billion to $9.0 billion. Of these loans exceed $1 million. The survey found that the loans originated by banks in that program in 2019, approximately 86 percent of banks with total assets less community banks originated approximately 46 percent. than $250 million responded that their C&I loans were Between 2012 and 2019, noncommunity banks saw their “almost exclusively” to small businesses, regardless of share of loan originations fluctuate, going from 62 percent the size of the underlying loans. For banks with total in 2012 to a high of 65 percent in 2015 and then dropping assets between $250 million and $1 billion, approximately to 54 percent in 2019, while the dollar amount dropped 75 percent of respondents stated that their C&I loans were from the 2015 high of $14.5 billion to $10.6 billion in 2019 “almost exclusively” to small businesses. On the basis of (Chart 4.11). these responses an adjustment to the share of C&I loans that are made to small businesses would have shown that Most important, Chart 4.12 shows that community small business loans as a percentage of total C&I loans at banks’ SBA loan originations support the assertion that banks with total assets less than $1 billion jumped from community banks do not limit their small business 56 percent of C&I loans to approximately 78 percent in 2019 loans to $1 million. Rather, as with the findings of the (or from $55 billion to $76 billion) (Chart 4.10).14 Therefore, Lending Survey, SBA data show that a majority of the loans the responses from the Lending Survey provide additional originated by community banks are for amounts greater support for the belief that community banks are lending than $1 million. to their local businesses despite the declines in share of small C&I loans and the slow C&I loan growth rates that are based solely on Call Report figures.

15 SBA 7(a) program loans provide 75 percent guarantees on working capital loans to small businesses in varying amounts up to $5 million . 14 The 2018 FDIC Small Business Lending Survey did not use the Loans are originated through a bank, credit union, or community community bank definition to differentiate between banks; rather, it development financial institution . The total amount approved during used the asset sizes of institutions . the fiscal year ending September 30, 2019, was $23 6 billion. .

4-10 FDIC Community Banking Study ■ December 2020 Chart 4.11 Small Business Administration 7(a) Loan Originations $ Billions 16 Community Bank SBA Originations Noncommunity Bank SBA Originations

14

12 54% 10 62% 46%

8

6 38%

4

2

0 2012 2013 2014 2015 2016 2017 2018 2019 Sources: Small Business Administration; FDIC. Notes: Represents only those SBA loans made by insured depository institutions. Percentages on bars represent share of total.

Chart 4.12

Small Business Administration 7(a) Loans by Dollar Size $ Billions <$100 Thousand 16 $100 Thousand - $250 Thousand 14 $250 Thousand - $500 Thousand $500 Thousand - $1 Million 12 >$1 Million

10

8

6

4

2

0 2012 2013 2014 2015 2016 2017 2018 2019 20122013201420152016201720182019 Community Banks Noncommunity Banks

Sources: Small Business Administration, FDIC.

Like Call Report data, SBA loan origination data show share in these groups has been declining, and community that community banks tend to make more—by count— banks are almost even in several categories. Community large SBA loans than small SBA loans, compared with banks’ share (by count) of loans originated for more than noncommunity banks, which tend to focus on smaller $1 million is almost equal to the share of loans originated SBA loans. As shown by Chart 4.13, noncommunity banks by noncommunity banks. This level is not surprising make the vast majority (80 percent)—by count—of because, as discussed above, community banks focus on loans below $100 thousand in value. This share has not loans that build relationships and may take more analysis changed since 2011. While noncommunity banks still and require an understanding of the business and the make the majority of loans in other size groups, their business owner.

FDIC Community Banking Study ■ December 2020 4-11 Chart 4.13 Small Business Administration 7(a) Loans by Count Percent by Number of Loans Originated 90 Community Banks Noncommunity Banks 80

70

60

50

40

30

20

10

0 >$1 Million $500 Thousand – $250 Thousand – $100 Thousand – <$100 Thousand $1 Million $500 Thousand $250 Thousand Sources: Small Business Administration; FDIC.

Small Business Lending: Summary of community banks’ C&I portfolios. Moreover, for such community banks, the share of small business loans in Analysis of Call Report data, of responses to the FDIC Small the C&I portfolio may compare favorably with the share of Business Lending Survey, and of SBA 7(a) loan origination small business loans in the portfolios of noncommunity data reveals that community banks continue to play a key banks. These local-minded banks focus on loans that role in providing funding that support small businesses. build relationships: the loans tend to be larger and more Despite declines in the numbers reported in Call Reports, hands-on, and they involve continued loan administration. data from both the FDIC Lending Survey and the SBA The evidence indicates, therefore, that community banks show not only that community banks make loans to small continue to be key supporters of small businesses in their businesses—loans often greater than $1 million—but local areas, and there is no reason to expect this support also that small business loans often represent a majority to decline.

Box 4.2 Small Business Lending and the COVID-19 Pandemic

The federal government’s first step in aiding small businesses was passage of the Coronavirus Aid, Relief, and Economic Security Act, which among other things provided $659 billion in funds for small businesses through the Paycheck Protection Program (PPP). The program is administered by the SBA and the U.S. Treasury, with applications for the funds submitted through banks, credit unions, Community Development Financial Institutions (CDFIs), and other financial institutions. The program was designed to provide an incentive for small businesses to keep their workers on the payroll during the initial weeks of the pandemic, when many states put stay-at-home orders into effect. The loan amounts were based on two months’ salary and employee expenses (January and February 2020). Loan terms included an interest rate of 1 percent, a two-year maturity that was extended to five years for loans originated after June 5, six months of loan payment deferral, and loan forgiveness if certain criteria are met.

As of August 8, 2020, over five million loans totaling more than $525 billion had been originated.a Like community banks’ share of the small business loans held by all banks, community banks’ share of PPP loans outstanding held by all banks was larger than their share of total C&I loans held by all banks. As of June 30, 2020, community banks held 13 percent of all banks’ C&I loans but more than 30 percent of PPP loans held at banks. Funding the PPP loans resulted in an annual C&I loan growth rate of 69 percent at community banks, compared with 16 percent at noncommunity banks.

continued on page 4-13

a U .S . Small Business Administration .

4-12 FDIC Community Banking Study ■ December 2020 Box 4.2, continued from page 4-12

While the PPP helped many small businesses initially, economic challenges related to the pandemic have continued to affect many small businesses. According to the American Bankruptcy Institute, commercial bankruptcy filings have increased 44 percent when comparing filings from April through September 2020 to the same time period in 2019. Additionally, according to Yelp Economic Average, more than 163,000 businesses have closed through August 31, 2020, from the start of the pandemic (March 1, 2020). The full effect of the pandemic on small businesses may not be fully known for several years.

Agricultural Lending Chart 4.14

In 2019, the more than two million farms in the United Distribution of U.S. Farm Sector Debt, 2019 States held nearly $419 billion in debt, with about 83 percent of that amount split evenly between commercial Total Farm Sector banks and the Farm Credit System (FCS) (Chart 4.14). Debt $419 Although the aggregate volume of dollars lent is nearly evenly divided between commercial banks and FCS institutions, the number of institutions that extend the Farm Credit Commercial All Other loans is vastly different. At year-end 2019, more than System Banks Lenders $178 $168 $72 4,300 banks (about 84 percent of all commercial banks) held agricultural loans, compared with the FCS network of 72 lending institutions. Source: USDA. Notes: Data are in billions of dollars. Sub-sectors do not add to total due to rounding. Rural communities rely on their community banks to fund agricultural production. As Chart 4.15 shows, at year-end 2019, although community banks held just of farm loans at commercial banks was approximately 12 percent of total banking industry assets, their share 70 percent.16

Chart 4.15

Distribution of Agriculture Loans Among FDIC-Insured Institutions, Year-End 2019 ($183 Billion) Number of Banks Agricultural Loans ($ Billions) by Bank Group by Bank Group

327 100 21.8 928 64.6

32.4

3,822

64.4 Community-Bank Agricultural Specialists Community-Bank Non-Agricultural Specialists 100 Largest Noncommunity Banks (NCBs) Other Noncommunity Banks Source: FDIC.

16 In 2019, community banks funded approximately 31 percent of farm sector debt .

FDIC Community Banking Study ■ December 2020 4-13 At year-end 2019, there were 928 community banks Community-Bank Agricultural Specialists Tend to that specialized in agricultural lending (“community- Be Small and Heavily Concentrated in the Center bank agricultural specialists”).17 They held 35 percent of the Country and to Have Large Exposures to of all agricultural loans held by commercial banks but Row Crop and Livestock Production represented only about 18 percent of all banks. The rest Community-bank agricultural specialists are typically of this section focuses on the performance and unique small, rural institutions. Remarkably, although as a characteristics of the community-bank agricultural group they hold about 35 percent of all agricultural loans, specialists. they hold just 1 percent of industry assets. The group’s median asset size is just $128 million, compared with Community-Bank Agricultural Specialists Performed Well Between 2012 and 2019 the nearly double $246 million for community-bank non-agricultural specialists. In fact, community-bank In the years leading up to and following the 2012 FDIC agricultural specialists tend to be the smallest of all Community Banking Study, the lending emphasis of community banks when the latter are grouped by lending community-bank agricultural specialists largely played specialty (Chart 4.16). More than 75 percent of the 928 in their favor. Their exposure to the negative credit effects community-bank agricultural specialists have total assets of the housing crisis and Great Recession was minimized, under $250 million, and just 19 have total assets in excess and instead they benefited from a strong, decade-long of $1 billion. farming boom. As shown in Map 4.1, 790 community-bank agricultural Starting in 2014 the agriculture sector struggled in terms of specialists, or 85 percent of the total, are concentrated profitability, but erosion in farm financial conditions was in just ten states in the center of the country. In 2019, gradual and generally modest in severity. Credit quality agricultural commodity receipts in these ten states totaled at community-bank agricultural specialists weakened $152 billion, or 41 percent of the $370 billion in total U.S. but still remained favorable by long-term historical agricultural commodity receipts. Agriculture in these ten comparison, and earnings and capital were strong. states is heavily focused on a handful of commodities:

Chart 4.16 Median Community Bank Asset Size by Lending Specialty Group, Year-End 2019 Median Asset Size $Millions 400 379 350 308 288 300 277 250 200 184 146 150 128 100 50 0 Ag (928) C&I (87) CRE (1,227) Cons (35) Multi (737)Mtge (731) None (1,005) Lending Specialty Group Designations Source: FDIC. Notes: Lending specialty groups are agricultural (Ag), commercial and industrial (C&I), commercial real estate (CRE), mortgage (Mtge), multi-specialty (Multi), consumer (Cons), and no specialty (None). Figures in parentheses denote number of community banks. Lending specialty group definitions can be found in Appendix A.

17 As shown in Appendix A, the FDIC defines community-bank agricultural specialists as community banks that have total loans greater than 33 percent of total assets and agricultural loans greater than 20 percent of total assets, and are not considered a multi-specialist .

4-14 FDIC Community Banking Study ■ December 2020 Map 4.1 Locations of Community-Bank Agricultural Specialists by Headquarters, Year-End 2019

Community-Bank Agricultural Specialist

Source: FDIC. Notes: Dot positions are based on locations of bank headquarters. The 10 shaded states contain the largest numbers of community-bank agricultural specialists by state. There are no community-bank agricultural specialists headquartered in Alaska or Hawaii.

cattle, corn, hogs, and soybeans. In 2019, within these Therefore, while community-bank agricultural specialists ten states, these four commodities totaled 77 percent of are exposed to nearly all types of agricultural production, total commodity receipts, and the ten-state aggregate they are most heavily exposed to a handful of row crops receipts of each of these commodities represented and livestock, with significantly less risk posed by other about two-thirds or more of total U.S. receipts for each agricultural production. commodity.18 These states are less concentrated in dairy and poultry production and far less concentrated in fruits, Agricultural Lending Is the Least Pervasive Lending nuts, and vegetable production. Segment Among Community Banks

Although the vast majority of community banks hold Conversely, areas in these ten states that are heavily at least some of each of the loan types constituting the concentrated in dairy, poultry, fruits and tree nuts, various loan specialist groups, if a particular loan segment and vegetables and melons are headquarters to few happens to be absent, it is most likely to be agriculture agricultural specialists. The seven states responsible for (Chart 4.17). A community bank is five times more likely to more than 90 percent of fruit and tree nut production have no agricultural loans than to have no C&I loans, and and three-quarters of vegetable and melon production 27 times more likely to have no agricultural loans than are headquarters to just 11 community-bank agricultural CRE loans. specialists.19

Moreover, there is far greater polarization of concentration in agricultural loan holdings than in CRE, 1–4 family residential mortgage, and C&I lending. As shown in 18 Aggregate receipts in the ten shaded states of Map 4 1,. as a percentage of total U .S . receipts, by commodity: cattle (65 percent), Chart 4.18, unless a bank holds sufficient agricultural loans corn (72 percent), hogs (73 percent), and soybeans (65 percent) . to warrant the label “agricultural specialist,” it tends to 19 The seven states represent the top five states in each commodity, hold agricultural loans in low proportion to its capital. The with overlap of some states . Similarly, there are just 34 community- bank agricultural specialists in the eight states whose leading only other lending specialty with similar polarization is commodity is dairy products (61 percent of U .S . production), and the consumer specialist group. there are just 34 community-bank agricultural specialists in the nine states whose leading commodity is commercial chickens (71 percent of U .S . production) .

FDIC Community Banking Study ■ December 2020 4-15 Chart 4.17 Chart 4.18 Shares of Community Banks Not Holding Loans Loan Type to Leverage Capital Ratios by Bank Group, Year-End 2019 by Loan Type, Year-End 2019 Median Ratio Share of Community Banks Holding Zero Dollar Balances of Given Loan Type Percent Community Bank - Specialty Group Community Bank - All Other Percent 400 16 14.9 Noncommunity Bank 350 14 300 12 250 10 200 8 150 6 100 4 2.9 1.7 50 2 0.5 0.8 0 0 Agricultural CRE Mortgage Consumer C&I Agricultural CRE Mortgage Consumer C&I Source: FDIC. Source: FDIC. Note: Lending specialty group definitions can be found in Appendix A.

Few New Banks Become Agricultural Specialists, headquartered in rural counties, and just 20 percent and the Group Is Dominated by Community Banks in metropolitan counties. That is the inverse of the That Have Historically Been Agricultural Specialists rural-urban mix of other community-bank loan specialist groups. One consequence of this inversion Although agricultural activity occurs just about is that community-bank agricultural specialists are everywhere in the United States, it is naturally most located in areas with vastly lower population densities, concentrated in rural areas, and as a result community as seen in Chart 4.19. Even when the focus is solely on agricultural specialists are also heavily concentrated in metropolitan areas, the average population density for rural areas (Chart 4.19).20 At year-end 2019, 57 percent agricultural specialists is still just 100 people per square of community-bank agricultural specialists were mile, suggesting that even when agricultural specialists

Chart 4.19 Shares of Community-Bank Loan Specialist Groups by County Urban Classification and Population Density, Year-End 2019 Headquartered County Description Average Population Density of Bank-Headquartered County Locations Metropolitan Area Micropolitan Area Rural Share of Bank Group Average Population Density Percent Persons per square mile 100 1,800 1,555 1,601 90 1,600 1,431 80 1,400 70 1,200 60 1,000 933 50 800 40 600 30 20 400 10 200 40 0 0 Agricultural Mortgage C&I Consumer CRE Agricultural Mortgage C&I Consumer CRE Sources: FDIC, U.S. Census Bureau. Notes: Figures were compiled using community banks as of year-end 2019, metropolitan and micropolitan delineation files as of March 31, 2020, and county populations based on the 2010 decennial census. Lending specialty group definitions can be found in Appendix A.

20 For purposes of this study, the FDIC has labeled all counties existing outside metropolitan statistical areas and micropolitan statistical areas as rural . This is consistent with the approach taken by FDIC authors in past studies on rural depopulation . See Anderlik and Walser (2004) and Anderlik and Cofer (2014) .

4-16 FDIC Community Banking Study ■ December 2020 are based in metropolitan areas, they tend to be based rural communities tend to reflect the characteristics of in smaller metros or in the less urban fringes of the their communities and are marked by generally slower metro areas. growth and high concentrations in agriculture. As a result, community-bank agricultural specialists tend to Moreover, half of agricultural specialists are remain true to their agricultural roots. Chart 4.20 shows headquartered in rural counties characterized by long- that 56 percent of the 793 community banks labeled as term population decline (see Box 3.1 in Chapter 3 for a agricultural specialists in 1990 continued to have the same more detailed analysis of such counties).21 These counties label in at least 28 of the subsequent 30 years. tend to have sparse populations, greater proportions of elderly people, and less-vibrant and less-diversified For the reasons discussed above, this tendency to remain economies than most other counties have.22 Such attached to their roots is most pronounced among conditions for the most part reflect the decades-long community-bank agricultural specialists headquartered consolidation in agriculture. Since these dynamics are in rural areas. Chart 4.21 shows this by juxtaposing the not conducive to new-bank formation, which largely pattern of community-bank agricultural specialists occurs in areas experiencing strong population and headquartered in growing metropolitan areas against economic growth, only 41 of the more than 1,400 new the pattern of agricultural specialists headquartered in banks formed since the beginning of 2000 were identified declining rural areas. Of the agricultural specialists in as an agricultural specialist either at formation or in any declining rural areas, 60 percent remained agricultural quarter since formation.23 specialists throughout the entire 30–year period 1990–2019, whereas the comparable rate for agricultural Meanwhile, absent branching into growing urban areas specialists based in growing metro areas was only or purchasing assets, community banks in declining 23 percent.

Chart 4.20

Shares of Community-Bank Agricultural Specialists by Number of Years Considered Agricultural Specialists, Year-End 2019 Percent of Institutions 60 55.5 50 40 30 20 7.6 10 3.8 3.9 4.7 3.7 5.2 5.8 4.7 5.3 0 1 to 3 4 to 6 7 to 9 10 to 12 13 to 15 16 to 18 19 to 21 22 to 24 25 to 27 28 to 30 Years Years Years Years Years Years Years Years Years Years Number of Years Categorized as an Agricultural Specialist 1990 Through 2019 Source: FDIC. Notes: Sample only includes banks open from January 1, 1990, through year-end 2019, that were considered a community bank during all quarters of 1990 and 2019, and were also considered an agricultural specialist in any quarter during 1990. Because of seasonality in agricultural lending, a bank is considered as having been an agricultural specialist in a given year if it was identified as an agricultural specialist in any quarter during that year.

21 Anderlik and Cofer (2014) . The FDIC defines counties as growing, declining, and accelerated declining on the basis of 30-year population trends . 22 Anderlik and Walser (2004) . 23 Of these banks, 1,130 were identified as community banks in the quarter in which they were established, and 302 as noncommunity banks .

FDIC Community Banking Study ■ December 2020 4-17 Chart 4.21

Share of Community-Bank Agricultural Specialists by County Urban Classification and Age, Year-End 2019 Share of Bank Group Percent 70 Growing Metro Counties 59.6 60 Declining Rural Counties 50 43.8 40 32.8 30 26.9 23.4 20 13.5 10 0 1 to 15 Years 16 to 29 Years30 Years Number of Years Categorized as an Agricultural Specialist 1990 Through 2019 Source: FDIC. Notes: Sample only includes banks open from January 1, 1990, through year-end 2019, that were considered a community bank at all quarters in 1990 and 2019, and were also considered an agricultural specialist in any quarter during 1990. Because of seasonality in agricultural lending, a bank is considered as having been an agricultural specialist in a given year if it was identified as an agricultural specialist in any quarter during that year. Metro counties are counties that are part of a metropolitan statistical area; rural counties are counties that are not part of a metropolitan statistical area or micropolitan statistical area. Growing counties had an increase in population between 1980 and 2010; declining counties had a decrease in population between 1980 and 2010.

Community-Bank Agricultural Specialists Remain As their annual growth in loan volume has demonstrated, Committed to Agricultural Lending Through community-bank agricultural specialists have been Agricultural Economic Cycles strongly committed to lending to producers through the peaks and valleys of agriculture operating returns In 2012, when the first FDIC Community Banking Study (Chart 4.22). In the period 2000–2019, they experienced was published, a decade-long boom in U.S. agriculture was only two quarters when aggregate agricultural production nearing its apex, buoyed by steep increases in commodity loan volume was lower than it had been in the same quarter prices and farmland values. At that time, farm financial one year earlier; those two quarters were fourth quarter conditions and community-bank agricultural credit quality 2016 (a decline of .08 percent from fourth quarter 2015) were as favorable as they had been in many decades. But and first quarter 2017 (a decline of 0.81 percent from first in the years after 2013, when farm incomes reached their quarter 2016). Never, however, did the group see a similar peak, the agricultural sector endured lower prices, weaker quarterly decline in aggregate farmland-secured loans. returns, and gradually deteriorating financial conditions.24 Noncommunity banks, on the other hand, demonstrated far Fortunately, most agricultural specialists maintained greater volatility in lending activity through the sector’s strong capital levels and loan loss reserves while peaks and valleys; in particular, they were far more prone simultaneously keeping in check their concentrations to pull back on agricultural loan volume as performance in farmland-secured lending. As a result, they had the weakened. The largest noncommunity banks saw strength and capacity to manage the rising stress in the production loan volumes decline in a total of 25 quarters farming sector, partly by cooperatively working with their between 2000 and 2019, and farmland-secured loan borrowers to restructure operating shortages using the volumes decline in 9 quarters. As Chart 4.22 shows, these borrowers’ strong equities in farmland.25 declines occurred often during dips in U.S. farm income.

Among community banks, although agricultural specialists and nonagricultural specialists showed similar 24 Reflective of the ongoing stress, the farm sector’s aggregate working capital balance declined by more than one-third from 2014 growth patterns in their agricultural lending and therefore to 2020, and its aggregate debt-to-asset ratio rose from 11 4. percent presumably similar commitment to the agriculture sector in 2013 to a forecasted 14 percent for 2020 . See U .S . Department of throughout the course of its ups-and-downs, from a risk Agriculture . Farm Balance Sheet and Financial Ratios, U .S . 25 This assertion is based on many anecdotal accounts reported perspective the nonspecialists tend to have far less at stake by examiners from the FDIC and other bank regulatory agencies, because of much smaller agricultural concentrations. agricultural bank officers, and representatives of industry trade groups .

4-18 FDIC Community Banking Study ■ December 2020 Chart 4.22 Year-Over-Year Growth in Agricultural Production and Farmland-Secured Loans, First Quarter 2000 Through Fourth Quarter 2019 Agricultural Production Loans Loans Secured by Farmland Year-Over-Year Change in Aggregate Balance Year-Over-Year Change in Aggregate Balance Percent All Other Noncommunity Banks Percent All Other Noncommunity Banks 35 Community-Bank Agricultural Specialists 35 Community-Bank Agricultural Specialists 30 Community-Bank Non-Agricultural Specialists 30 Community-Bank Non-Agricultural Specialists 100 Largest Noncommunity Banks 100 Largest Noncommunity Banks 25 25 20 20 15 15 10 10 5 5 0 -5 0 -5 -10 Shading Is Inflation-Adjusted Net Farm Income Shading Is Inflation-Adjusted Average Farm Real Estate -15 Scaled to Fit Y-Axis Maximum Value of 35 -10 Value Per Acre Scaled to Fit Y-Axis Maximum Value of 35 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: FDIC. Notes: Data are quarterly figures from March 31, 2000 through year-end 2019. The bank sample only includes institutions operating throughout the entire period, and grouping designations are based on group designation as of year-end 2019.

The big difference, however, is not between the two Box 4.3 Agricultural Lending and the community-bank groups—agricultural specialists and COVID-19 Pandemic nonagricultural specialists—but between both groups, on the one hand, and noncommunity banks, on the The COVID-19 pandemic disrupted the agricultural other hand. For regardless of exposure and risk in the sector, with reduced demand and mismatches community-bank sector, both groups are committed to and bottlenecks in the food-supply chain causing the farm sector through good times and bad. Meanwhile, commodity prices to fluctuate widely. COVID-19 noncommunity banks—and especially the largest, for outbreaks among workers caused temporary closures of dozens of large meat-processing plants in April and which agricultural lending is generally immaterial in May, which created a backlog of market-ready animals. proportion to their loan portfolios and capital—are prone These processing issues drove animal prices much to add and subtract credit exposure to the agricultural lower while at the same time drove meat prices higher sector as the sector’s performance outlook changes. for consumers. Closures of schools and restaurants cut demand for milk and dairy products, and some dairy Agricultural Lending: Summary farmers were forced to dump milk as a result. Crop and livestock prices fell sharply between March and Through their lending activities, community-bank June; prices have since rebounded to varying degrees. agricultural specialists are important to the nation’s Strong sales commitments from export countries for agricultural sector and rural communities. Although corn, soybeans, and pork have been positive news since representing a small percentage of all commercial banks mid-2020. and an even smaller percentage of industry assets, they provide more than one-third of all agricultural credit In December 2020, the U.S. Department of Agriculture funded by commercial banks. Agricultural specialists tend forecasted net farm income to increase from $84 billion to be small, yet by tending to the credit needs of many in 2019 to $120 billion in 2020. However, the forecast included a $24 billion, or 107 percent, increase in direct small and mid-sized farmers, they are a backbone of their federal farm payments to a record $46 billion. Most communities. Importantly, they are highly committed to of the increased assistance was pandemic-related. meeting those farmers’ credit needs even during periods The forecast also included a $5 billion reduction in of agricultural stress beyond their borrowers’ control. expenses. Without the added direct payments and lower Finally, by remaining committed to their agricultural roots, expenses, forecasted 2020 net farm income would be community-bank agricultural specialists keep banking much lower at $90 billion, but still 8 percent above alive in many rural areas whose demographic and economic 2019’s income level. profiles leave them unapproached by de novo activity.

FDIC Community Banking Study ■ December 2020 4-19

Chapter 5: Regulatory Change and Community Banks

The period 2008 through 2019 was one of intense In analyzing the effects of bank regulation on community regulatory activity, much of which affected community (or other) banks, it is important to recognize that the banks. So numerous were the new regulations that keeping conclusions reached are not definitive, given three current with them would have challenged any bank, but inherent challenges: decisions in banking are driven by especially a small bank with limited compliance resources. many factors other than regulation; community-bank Some of these regulatory actions created new obligations aggregates may mask behavioral responses within for banks, but many of them benefited banks. Some applied segments of the industry; and the goals of regulation only to specific classes of banks (such as national banks extend far beyond the effects on banks. For details on these or federal thrifts), many applied only to specific activities three challenges, see Box 5.1. or products, and some were technical clarifications or changes to the scope of various exemptions or exceptions. A common feature of these rules, however, is that the Box 5.1 Three Big Challenges to Pinning Down the Effects of Bank Regulation on Banks affected banks needed to understand them. Putting aside any consideration of the substantive effects of these The three most significant challenges to any attempt rule changes, their large number and scope make clear to determine the effects of bank regulation on banks of that merely being knowledgeable about changes in bank any size are as follows: regulation can be, by itself, an important and potentially First, bank decisions are driven by many factors other daunting task for any bank. than regulation. Those include decisions related to staffing and operations, the extent of involvement in Regulatory changes notwithstanding, community banks particular business lines, or even entry into or exit in aggregate have exhibited strong financial performance from the banking industry itself. The many factors since the crisis, as noted in Chapter 1 of this study, and besides regulation that bear on these decisions could include the state of loan demand, interest rates, or aggregate community bank loan growth has been strong. the ability to attract stable retail deposits; changes Yet as will be discussed in this chapter, the pace of in technology; changes in customer demographics; regulatory change and the volume of actions make plausible challenges with arranging for appropriate management the idea that some community banks, and particularly the succession; or consolidation of businesses in a bank’s smaller institutions among them, may have elected to exit market area.a particular business lines, or even the banking industry Second, community bank aggregates may mask itself, partly because of costs associated with regulatory behavioral responses within segments of the industry. compliance. The pace of regulatory change may have been For example, a particular type of lending may display one among a number of factors contributing to three post- a steady upward trend for community banks in the crisis developments: a high proportion (compared with aggregate, but a more complete picture might reveal other time periods and other banks) of small mortgage that regulatory developments caused some smaller lenders that reduced their residential mortgage holdings, community banks to exit that type of lending, with the the record rates at which community banks were exiting lending then migrating to larger community banks. the banking industry in the years leading up to 2019, and Another example might be an aggregate flat trend for noninterest expense, which might mean no increase an apparent increase in the target asset size of new small in regulatory compliance costs, or it might reflect banks as reflected in their initial equity. changes in bank behavior in response to regulation, with banks reallocating staff time or product mix, or Not included in the chapter is an analysis of the public adopting new technologies, to avoid an increase in policy goals of banking laws and regulations or how well noninterest expense. they have been achieved. Implicit to the presentation, continued on page 5-2 however, is the belief that a thriving community bank sector is worth preserving. If policy makers share that a For a discussion of how business consolidation may affect belief, bank regulation should achieve statutory goals in banks, see Brennecke, Jacewicz, and Pogach (2020) . a way that accommodates, to the extent appropriate, the business models of community banks.

FDIC Community Banking Study ■ December 2020 5-1 Chart 5.1 Box 5.1, continued from page 5-1 Selected Federal Regulatory Actions A third difficulty in quantifying the effects of bank Applicable to Community Banks regulation is that the goals of regulation extend far beyond the effects on banks. A partial list of statutory goals underlying the development of the U.S. bank regulatory framework includes promoting the financing of government activities, providing for a national currency, promoting a reliable payments system, ensuring sound and lawful bank operations, 12/12/07 12/12/09 12/12/11 12/12/1312/12/1512/12/17 12/12/19 promoting financial stability, protecting bank Source: Agency websites. depositors or other creditors while limiting the cost Note: Bars mark the announcement dates of 157 substantive final rules or of the federal banking safety net and determining federal programs aecting community banks that were issued by the FDIC, Federal Reserve, OCC, CFPB, Treasury, or FinCen. Rule changes depicted who bears that cost, protecting bank customers from include burden reducing rules and federal financial support programs unfair practices or illegal discrimination, combating benefitting banks. The chronology starts with the creation of the Federal Reserve’s Term Auction Facility in December 2007 and ends at year-end 2019. money laundering, avoiding monopoly or undue concentration, promoting lending, and supporting credit to underserved communities. Moreover, the very of regulatory changes. A summary follows, to be followed existence of a large body of bank regulation has given in turn by a brief discussion of regulatory changes that rise to the statutory and policy objective of simplifying have occurred as a result of the COVID-19 pandemic. An regulation and ensuring that it is appropriately tailored appendix—elaborated on in the next paragraph—extends to small, regulated entities. the chapter.

In short, not only do bank regulations have potentially Appendix B contains a chronology and a brief description wide-ranging effects outside the banking industry, but of selected federal rules and programs that applied to the narrower effects on banks themselves can be difficult community banks and were put in place from late December to pin down. This suggests that gaining perspective on 2007 to year-end 2019. The chronology is limited almost banking trends requires a holistic perspective. The FDIC entirely to substantive final rules and federal programs of conducts a significant amount of banker outreach, meets the FDIC, Board of Governors of the Federal Reserve System regularly with its Community Bank Advisory Committee, (Federal Reserve), Office of the Comptroller of the Currency, and benefits from public comments on its rules, including Consumer Financial Protection Bureau, and the Department those received as part of the Economic Growth and of the Treasury, including rules of the Financial Crimes Regulatory Paperwork Reduction Act process. Given the Enforcement Network (FinCEN). The appendix generally important challenges and caveats associated with the does not include the following: Call Report changes; analysis, this chapter should be viewed as part of an changes to accounting standards; tax changes; supervisory ongoing dialogue about community bank regulation and guidance; statements of policy; changes in state laws or not as a source of firm and final conclusions. regulations; ministerial rules such as inflation adjustments, rules issued in connection with changes in regulatory The remainder of the chapter begins with a brief review authority from one agency to another, or technical changes of the level and trend of noninterest expense ratios at to agency procedures; or rules that apply exclusively to community banks, since that category would typically large or internationally active banks. Rules issued by include direct expenses associated with regulatory multiple agencies, and rules issued as both interim final compliance. That review is followed by an overview of and final, are counted only once. the major changes to federal regulations and programs affecting community banks, starting with the three broad Even with these restrictions, the appendix lists 157 final categories of rules and programs most directly tied to the rules and programs applying to community banks, an 2008–2013 banking crisis: deposit insurance and other average of 1 every 28 days during the 2008–2019 period federal financial dealings with banks, capital adequacy (Chart 5.1).1 rules, and residential mortgage and servicing rules. The chapter continues with observations about community- 1 Rules finalized after 2019 are not covered in this chapter or its bank exit and entry as possible indicators of overall effects appendix, apart from a reference in a concluding text box to selected pandemic-related regulatory actions taken in 2020 .

5-2 FDIC Community Banking Study ■ December 2020 Chart 5.2

Noninterest Expense of Community and Noncommunity Banks Noninterest Expense to Average Assets (Percent) 4.00 Community Banks Noncommunity Banks 3.50

3.00

2.50

2.00

1.50 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: FDIC. Note: Full year noninterest expense as percent of average trailing 5-quarters assets, 1996–2019. Gray bars denote recession periods.

Noninterest Expense Is Highest at Small costs, so that the effects of the regulatory change may Community Banks not be evident in noninterest expense. Finally, banks may incur regulatory compliance costs that do not show up in The assessment of the effects of regulatory changes noninterest expense. For example, bank staff time devoted will benefit from a preliminary examination of trends to compliance may divert time from other strategic or in noninterest expense. Noninterest expense includes revenue-generating activities. expenses for salary, premises, legal and consulting fees, information technology (including ensuring the security Chart 5.2 depicts the trend in noninterest expense ratios at of that technology), and a variety of other noninterest community versus noncommunity banks. The chart shows expenses. Direct expenses associated with regulatory that notwithstanding the regulatory developments since compliance often fall within this category, and therefore 2008, community banks’ aggregate noninterest expense changes in, or levels of, noninterest expense relative to ratios declined modestly. The chart also shows that banks’ overall revenue and cost structures may provide community banks’ noninterest expense ratios have been indirect evidence of regulatory effects. slower to decline than those of noncommunity banks. This may reflect a community bank business model involving There are four important caveats to the discussion of more direct interaction with customers, in addition to noninterest expense. First, much of noninterest expense fixed costs that are a higher percentage of small banks’ would be necessary to conduct a banking business even cost-structures given their smaller asset size. Both of in the absence of regulation, so changes in the level and these factors may impose practical limits on how much trend of noninterest expenses may reflect changes in the noninterest expense ratios can be reduced. Thus, to the way banks do business that are unrelated to regulation. extent that there was an increase in expense arising from Second, the portion of noninterest expense attributable regulatory change, the effect may be greater relative to the to regulatory compliance is unknown to researchers, and overall cost structure of a typical community bank than to even bankers may have difficulty estimating these costs.2 that of a large noncommunity bank. Third, as noted above, banks may respond to changes in regulation by changing their behavior to avoid regulatory As indicated in Chart 5.3, community banks are not a 2 Call Reports include line items for legal expense, consultant homogenous group with respect to their noninterest expense, and accounting and auditing expense, but reporting expense ratios. Smaller community banks have had thresholds are such that many small banks need not report these items, and just as with other noninterest expenses, it is not possible substantially higher noninterest expense ratios than to determine the portion of these expenses that banks would need to larger community banks. Noninterest expense ratios at incur even in the absence of regulation .

FDIC Community Banking Study ■ December 2020 5-3 Chart 5.3 Community Bank Noninterest Expense by Asset Size Noninterest Expense to Average Assets (Percent) 3.60

3.40

3.20

3.00

2.80

2.60 Assets ($Millions): < $100 2.40 $100 – $249 $250 – $499 2.20 $500 + 2.00 1996 1998 2000 2002 2004 2006 200820102012201420162018 Source: FDIC. Note: Full year noninterest expense to average 5-quarter trailing assets, 1996–2019. Gray bars denote recession periods.

Chart 5.4

Community Bank Profitability by Asset Size Pretax Return on Average Assets Assets ($Millions): (Percent) < $100 2.00 $100 – $249 $250 – $499 $500 + 1.50

1.00

0.50

0.00

-0.50

-1.00 1996 1998 2000 2002 2004 2006 200820102012201420162018 Source: FDIC. Note: Full year pretax net income as percent of average 5-quarter trailing assets, 1996–2019. Gray bars denote recession periods.

community banks with assets less than $100 million, with assets above $500 million, while small banks’ pretax which at year-end 2019 constituted about 24 percent of all return on assets was 53 basis points lower (Chart 5.4). community banks, averaged 48 basis points higher during the years 1996–2019 than for community banks with Charts 5.3 and 5.4 indicate that higher overhead and lower assets greater than $500 million.3 Smaller banks’ higher profitability at smaller community banks are not new expense ratios have a substantial negative effect on these developments of the post-crisis period. The charts make banks’ profitability: in 2019, noninterest expense ratios at clear, however, that if higher noninterest expenses were community banks with assets less than $100 million were the outcome of a regulatory change, that cost would weigh 47 basis points more than the ratios at community banks relatively more heavily on smaller banks. For example, in considering the profitability effects of a hypothetical 3 Income and expense items in basis points are relative to average (emphasis added) increase in bank staff that generates no assets .

5-4 FDIC Community Banking Study ■ December 2020 additional revenue, Feldman, Heinecke, and Schmidt (2013) A subsequent important change in deposit insurance estimated that “the median reduction in profitability for arrangements was the statutory change in the assessment banks with less than $50 million in assets is 14 basis points base from domestic deposits to assets minus tangible if they have to increase staff by one half of a person.” equity capital. Since large banks tend to obtain a greater proportion of their funding from non-deposit sources None of this information bears on either the core than do small banks, the change in the assessment base profitability of community banks, or the variation over shifted some of the cost of deposit insurance assessments time in community bank profitability caused by economic from small banks to large banks. For second quarter 2011, factors, as discussed, for example in Fronk (2016). Instead, when the changes to the assessment base became effective, the discussion here highlights that the profitability of assessments for banks with less than $10 billion in assets community banks in general, and smaller community were 33 percent lower in the aggregate than first quarter banks in particular, reflects a higher proportion of assessments, and those banks’ share of total assessments noninterest expense in their cost structures and, given decreased from about 30 percent to about 20 percent. their smaller asset size, a greater sensitivity of profitability to any given increment of noninterest expense, including The allocation of the cost of building and maintaining the an increment to expense that might be necessary as a Deposit Insurance Fund (DIF) changed in other ways. The result of a change in regulation. Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank Act) increased the minimum reserve Deposit Insurance and Other Federal Financial ratio of the fund from 1.15 percent to 1.35 percent, required Dealings With Banks Changed in Important that the reserve ratio reach that level by September Ways as a Result of the Financial Crisis 30, 2020, and required that the FDIC offset the effect of the increase on small banks. To implement these The regulatory actions that were the most immediate requirements, the FDIC imposed surcharges on large response to the 2008 financial crisis were those pertaining banks, generally those with assets greater than $10 billion. to the federal banking safety net that supported banks As of September 30, 2018, the reserve ratio exceeded the during the crisis, and that in some cases permanently required minimum of 1.35 percent, and the surcharges benefited small banks relative to large banks. Many were suspended. Furthermore, to implement the Dodd- community banks borrowed from the Federal Reserve’s Frank Act requirement that the FDIC offset the effect of the Term Auction Facility (TAF), in which the Federal Reserve increase on small banks, the FDIC awarded $765 million lent to banks against a broader range of collateral than in assessment credits to small banks for the portion of was accepted at the Discount Window. Many community their regular assessments that contributed to growth in banks also participated in the U.S. Treasury’s Capital the reserve ratio between 1.15 percent and 1.35 percent. The Purchase Program (CPP), in which the Treasury invested FDIC remitted the final remaining assessment credits to in subordinated debt or preferred stock of viable banks small banks on September 30, 2020. The FDIC also made and bank holding companies. In addition, community significant changes in deposit insurance pricing intended banks benefitted from the 2008 temporary increase in to more accurately reflect risk, so that a less risky bank the standard deposit insurance limit to $250,000 (which does not subsidize activities of a riskier bank that could was made permanent in 2010), and from the FDIC’s increase loss to the DIF. These changes were not statutorily Temporary Liquidity Guarantee Program (TLGP), whose required but reflected the FDIC’s historical experience with two components were guarantees of holding company the risk characteristics of failed banks. obligations, and temporary unlimited deposit insurance

4 coverage of noninterest-bearing transaction accounts. The Federal Reserve also made important changes in its financial dealings with banks. The Federal Reserve announced in October 2008 that it would begin to pay interest on depository institutions’ required and excess 4 A list of debt issuances guaranteed by the FDIC during the crisis pursuant to the Temporary Liquidity Guarantee Program can be found reserve balances. In 2016, the Federal Reserve implemented at https://www fdic. gov/regulations/resources/tlgp/total_debt. html. . a statutory requirement by reducing the dividend paid to The amount of noninterest-bearing transaction accounts guaranteed by the FDIC for the institutions that opted in to the Transaction large banks (with assets greater than $10 billion) on their Account Guarantee program can be found on Call Report schedule Federal Reserve bank stock from 6 percent, to the lesser RC-O, memorandum item 4 .

FDIC Community Banking Study ■ December 2020 5-5 of 6 percent or the most recent ten-year Treasury auction them. The rule and its associated statute were intended rate before the dividend, while smaller banks’ dividend to relieve extremely well-capitalized banks of the burden rate remained at 6 percent. This latter change affects only of calculating risk-based capital requirements. As of first banks that are members of the Federal Reserve System. quarter 2020, slightly less than 40 percent of the 4,327 eligible banks in the United States had chosen to adopt the Changes in Capital Regulation Were Mainly but community-bank leverage framework. not Only About Implementation of Basel III Under Basel III, Community Banks Built Capital The most important change to capital adequacy regulation More Than Noncommunity Banks, and Grew Their during the 2008–2019 period was U.S. implementation of a Loans Faster as Well version of the Basel III capital framework. However, other The U.S. banking agencies proposed the Basel III rule in important regulatory capital changes occurred during those 2012 and finalized it in 2013, with an effective date for most years, including temporary capital relief measures during banks of January 1, 2015, and a phase-in period scheduled the 2008 financial crisis and risk-based capital changes to end January 1, 2019 (year-end 2012 through year-end implemented in response to a change in the accounting 2018 is referred to here as the Basel III response period). for certain securitized assets. Another important change Broadly speaking, the new rules (1) increased the numerical was the statutory increase in the asset size threshold for level of risk-based capital requirements by 2 percentage the Federal Reserve’s Small Bank Holding Company Policy points while leaving leverage requirements for most FDIC- Statement, from $500 million to $1 billion and then again insured institutions unchanged; (2) changed certain risk to $3 billion. Bank and thrift holding companies subject weights; and (3) restricted the recognition in regulatory to that policy statement are not subject to consolidated capital of certain assets, and of certain debt instruments leverage- or risk-based capital requirements.5 A 2019 (Trust Preferred Securities) that were formerly included in rule implemented a statutory requirement to allow regulatory capital for bank holding companies. qualifying banks to opt in to a community bank leverage ratio framework, in which they are exempt from risk- As background, banks must maintain capital at a specified based capital requirements if they operate subject to a minimum ratio of their assets. For community banks, this higher leverage requirement than otherwise applies to simple leverage ratio requirement was not changed by Basel III.

Chart 5.5

Leverage Capital Ratios of Community and Noncommunity Banks Tier 1 Capital to Average Assets (Leverage Ratio) Percent 12

10

8 Community Banks 6 Noncommunity Banks

4

2

0 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: FDIC. Note: Tier 1 capital at year-end as percent of assets for the leverage ratio. Gray bars denote recession periods.

5 The Small Bank Holding Company Policy Statement does contain exceptions whereby some bank holding companies with assets less than the size threshold may be subject to consolidated capital requirements .

5-6 FDIC Community Banking Study ■ December 2020 Chart 5.6

Dividends and Capital Ratios of Community and Noncommunity Banks Percent Percent 25 80 Noncommunity Bank Dividend Payout Ratio (Right) 70 20 60 Community Bank Dividend Payout Ratio (Right) 50 15 Leverage Ratios: Community Bank Noncommunity Bank 40

10 30

20 5 10

0 0 2012 2013 2014 2015201620172018 Source: FDIC. Note: Leverage ratio is tier 1 capital to average assets. Dividend payout ratio is full-year dividends on common stock as percent of full year net income.

Throughout 2012–2019, community banks also had to Chart 5.7 ensure their capital exceeded specified ratios of so-called Growth of Community and Noncommunity risk-weighted assets. Basel III increased the required Bank Loans and of GDP risk-based ratios and changed some of the methods for Annual Percent Merger-Adjusted Growth Annual Percent Growth in: calculating risk-weighted assets, and as a result, many 10 Community Bank Loans Noncommunity Bank Loans banks held more capital. Because of its simplicity, the 9 Nominal GDP 8 leverage ratio is the easiest way to describe how much 7 capital banks hold, and it is used throughout this analysis to 6 describe capital trends during the Basel III response period. 5 4 At year-end 2019, both community banks and 3 2 noncommunity banks had leverage ratios higher than at 1 any time since data were reported in this format, and about 0 2 percentage points higher than their banking crisis lows 2012 2013 2014 2015 2016 2017 2018 2019 Sources: FDIC and Bureau of Economic Analysis (Haver Analytics). (Chart 5.5). Some of the increase in leverage ratios depicted in the chart is likely attributable to banks’ rebuilding capital from the losses experienced in the crisis, and some noncommunity banks, they also grew their loans on a is likely attributable to Basel III. merger-adjusted basis faster than noncommunity banks and faster than nominal GDP. Charts 5.6 and 5.7 illustrate Chart 5.5 shows that during the Basel III response period, the important point that higher or increasing capital ratios community banks had higher leverage ratios, and do not automatically imply lower loan growth, because increased those ratios more, than did noncommunity banks can increase their capital ratios by growing capital banks. Chart 5.6 shows that dividend policies were an rather than by reducing loan growth. important driver of these trends. From 2013 forward, New Basel III Regulatory Capital Deductions Did Not community banks’ dividend payout ratios never exceeded Affect Most Community Banks 50 percent. The payout ratios of noncommunity banks were never less than 60 percent, partly explaining why With these broad comparisons to noncommunity banks for noncommunity banks’ leverage ratios remained at least context, we now turn to a more specific discussion of Basel a full percentage point less than the comparable ratios of III relative to community banks. As indicated in Table 5.1, community banks. Chart 5.7 shows that during 2012–2018, Basel III was proposed in 2012, published as a final rule in while community banks grew their capital more than 2013, and phased in for community banks from January 1,

FDIC Community Banking Study ■ December 2020 5-7 Table 5.1 Median Leverage Ratios of Community Banks, 2012–2018

Date (Year End of Each Year) 2012 2013 2014 2015 2016 2017 2018 Leverage Ratio (Percent) 9 .90 10 .12 10 .25 10 .40 10 .46 10 .54 10 .87 Abbreviated Basel III chronology for banks not subject to the advanced approaches: August 2012: proposed rule July 2013: final rule published January 1, 2014–December 31, 2014: old rule in effect January 1, 2015–January 1, 2019: Basel III phase-in period* *Certain originally scheduled deductions from regulatory capital were subsequently eliminated . Source: FDIC .

Chart 5.8 Basel III Threshold Deductions of Community Banks Percent of Banks and Assets by Capital Deduction Range 100

80 Percent of Community Banks Percent of Community Bank Assets 60

40

20

0 Zero < 11 – 4.9 5 – 9.9> 10 Highest Deduction as Percent of Tier 1 Capital Number 4,699 423 406 137 109 Assets $1,690 $301 $179 $78 $43

Source: FDIC. Note: Data are for community banks, 2015–2019, assets in billions. Each bank’s maximum threshold deduction as a percent of tier 1 capital for any year end is tabulated. Assets are the bank's average assets over 2015–2019.

2015 to January 1, 2019. Table 5.1 tracks the year-by-year introduction of these deductions from regulatory capital, median leverage ratios of community banks during this known as “threshold deductions.” As Chart 5.8 indicates, time. Roughly speaking, by year-end 2018 the median these deductions did not affect most community banks: community bank was operating with $11 in tier 1 capital 80 percent of community banks never had a threshold per $100 in assets, up from $10 per $100 in 2012. Historical deduction in any year-end through 2019. The chart also experience has been that banks with more capital have indicates that the deductions were material for some lower failure rates, as discussed, for example, in Crisis and institutions, amounting (for example) to more than 6 Response: An FDIC History, 2008–2013. All else equal, this 10 percent of tier 1 capital for 109 institutions at some point aspect of Basel III should make community banks more during the years 2015–2019. resilient in periods of stress. Healthy Community Banks Increased Capital Ratios In addition to requiring a higher level of regulatory capital, by Retaining Earnings and Raising Capital, Basel III tightened limits on the capital recognition of While Weaker Banks Were More Likely to Curtail deferred tax assets and mortgage servicing assets, and Loan Growth introduced limits on the recognition of investments in the It is interesting to know how community banks effected capital of other financial institutions.7 An important part the increase in capital ratios during the Basel III response of the phase-in referenced in Table 5.1 was the gradual period. Broadly speaking, a bank that increases its capital ratios must increase its capital by a larger percentage 6 See page 123 of Crisis and Response . amount than it increases its loans or other assets. Some 7 Mortgage servicing activity of community banks is discussed in the banks might do this by maintaining growth in their loans next section of this chapter .

5-8 FDIC Community Banking Study ■ December 2020 Table 5.2 Components of Community Bank Capital Ratio Changes, 2012–2018

6-Year Total (Percent of 2018 6-Year Growth End of End of 2012 Leverage Assets) (Percent) 2018 PDNA Leverage (Percent Inflow: Inflow: Outflow: Leverage Community Assets Ratio of Tier 1 Net Capital Common Tier 1 Leverage Ratio Banks That Were: Number ($ Billions) (Percent) Capital) Income Raise Dividends Capital Assets Loans (Percent) Less Than Well Capitalized 97 28 5 .04 210 2 .44 4 .33 -0 .60 157 26 40 10 .23 Well Capitalized With: Low PDNA 2,992 767 10 70. 8 6 .36 0 .75 -2 .97 48 40 62 11 .25 Medium PDNA 878 251 10 50. 30 5 .47 0 .55 -2 .63 36 28 43 11 .16 High PDNA 339 102 9 .03 87 4 .00 1 .44 -2 .11 40 7 20 11 .87 Well Capitalized Low PDNA With: High RBC 2,631 652 11 04. 8 6 .38 0 .65 -2 .97 44 38 61 11 .54 Med RBC 323 107 8 .75 9 6 .25 1 .20 -2 .95 71 54 65 9 .73 Low RBC 38 8 8 .25 5 6 .32 1 .74 -2 .81 96 66 72 9 .71 Source: FDIC . Note: Only includes community banks reporting at every year end from 2012 through 2018 that made no acquisitions . “Leverage assets” refers to “total assets for the leverage ratio” from Call Report schedule RC-R . Well capitalized banks are grouped in two ways . Past due and nonaccrual (PDNA) loan ratio - defined as 90 days past due, nonaccrual, and other real estate owned - grouped by less than 20 percent of tier 1 capital (Low PDNA), 20 percent to 50 percent of tier 1 capital (Med PDNA), and greater than 50 percent of tier 1 capital (High PDNA) . High risk-based capital (High RBC) is tier 1 capital to risk-weighted assets greater than 12 percent, medium RBC (Med RBC) is tier 1 capital to risk-weighted assets between 10 and 12 percent, and low RBC (Low RBC) is tier 1 capital to risk-weighted assets below 10 percent . While net income, external capital raises, and outflows in the form of dividends on common stock are important factors explaining the change in equity capital from one time period to the next, they are not the only factors . The three inflow and outflow columns in this table are not intended to permit a complete reconciliation of the change in capital ratios from 2012 to 2018 . or other assets while retaining more of their earnings earnings available for retention. These banks generally had or raising capital externally, while others might not be a greater need than other banks to increase their capital willing or able to increase their capital but instead might ratios. They did so with a combination of more substantial grow their loans or other assets more slowly. Table 5.2 capital raises and slower loan growth than other banks. details the drivers of changes in capital ratios during 2012– 2018 for 4,306 institutions that were community banks In more detail, Table 5.2 depicts the 2012 and 2018 leverage in 2012, that reported in every year-end through 2018, ratios of various groups of banks, along with the inflows and that did not acquire another bank. Analyzing trends during the full six years from income and capital raises, for this population of banks allows a focus on the capital and the outflows from dividends, expressed in units of management decisions of banks in continuous existence the 2018 leverage ratio. Thus, for example, the 97 banks during the Basel III response period. in row 1, which were less than well capitalized in 2012 under the old rules, increased their leverage ratios from Before discussing Table 5.2 in any detail, it may be helpful about 5 percent to about 10.2 percent during the six years, to summarize the conclusions it appears to suggest. While with about 4.3 percentage points of the 5.2 percentage community banks as a group increased capital in response point increase contributed by capital raises. Another way to Basel III, healthy community banks do not appear to of accounting for the increase in leverage ratios is that it have curtailed loan growth in order to do this. For healthy reflects faster growth of capital than of assets, and these banks, even those with relatively lower initial capital, growth rates, along with that of loans, are also reported. earnings retention and capital raises were sufficient to Thus, for example, these 97 banks grew their loans increase capital ratios while maintaining strong loan 40 percent during the six years and their assets 26 percent, growth. Banks with higher levels of troubled assets or but roughly doubled their leverage ratios because their that were less than well capitalized, generally had lower capital increased by 157 percent.

FDIC Community Banking Study ■ December 2020 5-9 The next three rows segment the 4,209 well-capitalized they did so with earnings retention and comparatively community banks by the ratio of noncurrent loans and higher capital raises, while maintaining higher rates of leases plus other real estate to assets as of year-end 2012. loan growth than any other subset of banks considered in The two groups with higher levels of troubled assets the table. started the period with lower leverage ratios, earned less income over the period, and grew their leverage ratios Institutions Resulting From Community Bank Mergers through a combination of higher capital raises, somewhat Generally Had Lower Capital Ratios Than Before lower dividends, and somewhat slower loan growth. Some the Mergers of the banks in these two groups may have been subject Table 5.3 provides information about the capital effects to supervisory directives to limit growth at some point of mergers during acquisition years. The table shows during the six years. that acquirers generally had lower leverage ratios than the banks they acquired, especially toward the end of The last three rows limit the focus to 2,992 well- the 2012–2019 period; that acquirers raised capital and capitalized community banks with low levels of troubled paid dividends at rates that exceeded community bank assets. Their approaches to capital management during the averages during acquisition years; and that on a merger- six years were more likely to reflect “pure” responses to adjusted basis, leverage ratios of the resulting entities Basel III, without a separate motive to build capital coming were typically lower than before the acquisition. Higher from high volumes of troubled assets or supervisory dividends and capital raises may reflect anticipated directives. The table segments these generally healthy merger-related benefits such as those derived from banks by their initial tier 1 risk-based capital ratios. Banks eliminating duplicative overhead costs over time. With in the low and medium capital groups were those that had regard to the reduction in leverage ratios, it is possible chosen to manage to lower capital ratios, but then may that acquirers tended to have greater focus on growth have had an impetus from Basel III to increase those ratios and return-on-equity than did the non-acquiring banks in order to maintain what they viewed as an appropriate depicted in Table 5.2. Whatever the reason, the effects cushion above the new Basel III requirements. The of acquisitions on community bank leverage ratios ran importance of the last three rows is that while the banks directionally counter, albeit modestly, to the general in the two lower capital groups did increase their leverage increase in leverage ratios reported in Table 5.2. ratios more than the banks in the higher capital group,

Table 5.3 Leverage Ratios, Capital Ratios, and Dividends in Community Bank Mergers, 2013–2019

CBs Acquiring During Year CBs Acquired During Year One Year Acquiring Year-Ago Change in Banks’ Leverage Leverage Leverage Leverage Leverage Leverage Ratio Acquiring Average Assets Ratio Assets Ratio Ratio (Merger- Banks’ CB Average Dividend CB Average (Billions, (Percent, (Billions, (Percent, (Merger- Adjusted, Capital Raise Capital Raise Payout Dividend as of Prior as of Prior as of Prior as of Prior Adjusted Percentage (Percentage (Percentage Ratio Payout Ratio Year Number Year End) Year End) Number Year End) Year End) Percent) Points) Points) Points) (Percent) (Percent) 2013 146 $95 9 .46 164 $31 9 .67 9 .51 0 .01 0 .42 0 .17 57 50 2014 166 $137 10 .20 186 $39 10 .00 10 .15 -0 .16 0 .46 0 .19 83 49 2015 196 $198 10 .05 219 $46 10 .40 10 .12 -0 .31 0 .45 0 .18 65 50 2016 191 $194 10 .38 204 $47 10 .28 10 .36 -0 .24 0 .61 0 .22 70 50 2017 146 $151 10 .28 169 $47 10 .60 10 .36 -0 .07 0 .73 0 .32 76 47 2018 178 $254 10 .07 201 $55 10 .89 10 .22 -0 .05 0 .75 0 .25 54 45 2019 157 $220 10 .55 171 $53 12 .02 10 .84 -0 .60 0 .69 0 .29 81 51 Source: FDIC . Note: CB = Community Bank . Leverage assets is “total assets for the leverage ratio” from Call Report schedule RC-R . Change in leverage ratio is the difference from the prior year (for example, in the last row, -0 .60 signifies that the year end 2019 leverage ratio for the acquiring banks was 10 .24 percent) . Capital raise is sum of net sale of stock and other transactions with stockholders, in percentage points of leverage assets as of the year end for the row . Dividend Payout Ratio is dividends on common stock as a percent of net income during acquisition year . Table includes affiliated and unaffiliated acquisitions but no failed bank acquisitions . For this table, a community bank is a bank that meets the community bank definition at any of the year ends from 2013 to 2019 .

5-10 FDIC Community Banking Study ■ December 2020 Many Important New Regulations Dealt of servicers that were depository institutions serviced With 1–4 Family Residential Mortgage 5,000 or fewer loans.8 Lending and Servicing Community Banks’ Mortgage Growth Has Outpaced Between July 2008 and November 2019, largely in response Growth of U.S. Mortgages Outstanding and Growth of to laws enacted to address abuses in subprime and Mortgages of Noncommunity Banks alternative residential mortgage lending and mortgage servicing, federal agencies issued 36 distinct substantive Community bank mortgage lending since the banking final rules governing various aspects of 1–4 family crisis needs to be considered in the context of broader residential mortgage lending and mortgage servicing mortgage trends. First, the bursting of the pre-crisis (in this chapter, any reference to “mortgages” refers housing bubble left an imprint in the data that still to 1–4 family residential mortgages). The peak of this existed at year-end 2019: the total volume of outstanding rule-writing activity occurred in January 2013, when the 1–4 family residential mortgages in the United States Consumer Financial Protection Bureau (CFPB) issued six declined for seven years starting in 2008 and, while slowly substantive final rules (five alone and one jointly with recovering, as of year-end 2019 it remained just below other agencies) addressing residential mortgage lending the 2008 peak of $11.3 trillion. Second, at year-end 2019 and servicing. Changes to the residential mortgage and the housing government-sponsored enterprises (GSEs) mortgage servicing rules, based on their sheer number and GSE mortgage pools held 63 percent of outstanding and scope, have a strong claim to being viewed as the most U.S. 1–4 family residential mortgages, a historic high. It is important of the post-crisis regulatory changes. possible that the Qualified Mortgage safe harbor for loans sold to GSEs contributed to the growth of GSE holdings. Broadly and collectively, the mortgage rules addressed Third, at least among the largest originators and servicers matters including but not limited to: (1) establishing of 1–4 family residential mortgages, the share of nonbank 9 disclosure, registration, and qualification standards firms increased in the years before 2019. for mortgage loan originators, and the bases on which mortgage originators could be compensated; (2) defining Despite the generally subdued backdrop for aggregate high-cost mortgages and capping or prohibiting certain residential mortgage lending during the post-crisis period, fees and loan terms for them, and requiring borrowers and notwithstanding the new regulations, community for those mortgages to receive housing counseling; banks as a group continued to grow their residential (3) establishing ability to repay standards with which a mortgage portfolios. As of year-end 2019, over 99 percent defined class of Qualified Mortgages was presumed to of community banks reported some level of 1–4 family comply; (4) requiring appraisals, including—for certain residential mortgages, a percentage that has held steady higher-priced mortgages—a physical inspection of the for many years. As Chart 5.9 shows, between 2011 and 2019, interior of the property; (5) excepting small rural lenders the dollar weighted average mortgage loan to asset ratio of from certain requirements; and (6) providing that, on a community banks held steady at about 20 percent and was time-limited basis, mortgages sold to the federal housing only slightly down from its 2005 level of 22 percent shortly enterprises were deemed Qualified Mortgages. before housing prices reached their pre-crisis peak. This steady trend contrasts sharply with the decline in the same The servicing rules, among other things: (1) prohibited ratio for noncommunity banks. And, notably, between a number of specific mortgage servicing practices; 2012 and 2019, the merger-adjusted growth of residential (2) prohibited foreclosures while an application for a mortgage loans on the balance sheet at community banks mortgage modification was under review; (3) required far exceeded the merger-adjusted growth of mortgage servicers to inform borrowers who missed two consecutive loans of noncommunity banks and the overall growth of payments about loss-mitigation options to retain U.S. 1–4 family residential mortgage loans outstanding their homes; and (4) included exceptions from certain (Chart 5.10). requirements for servicers that service 5,000 or fewer loans that they or an affiliate originated. For context regarding the importance of the small servicer exemption, 8 CFPB (2019) estimated that as of year-end 2015, 95 percent See page 106 of CFPB (2019) . 9 Shoemaker (2019) .

FDIC Community Banking Study ■ December 2020 5-11 Chart 5.9

Residential Mortgages of Community and Noncommunity Banks Percent of Assets 35

30

25 Community Bank 20

15

Noncommunity Bank 10

5

0 1984 1988 1992 1996 20002004200820122016 Source: FDIC. Note: 1–4 family residential mortgages at year end as percent of assets at year end. Gray bars denote recession periods.

Chart 5.10 Chart 5.11 Growth of Community and Noncommunity Bank Mortgages Residential Mortgage Servicing by Community Banks and of U.S. Mortgages Percent of Banks and Assets 35 Annual Percent Growth Community Banks 8 Noncommunity Banks 30 All U.S. Outstanding 6 25 4 20 2 15 Community Bank Mortgage Servicing: 0 10 Percent Servicing Mortgages -2 5 Serviced Mortgages as Percent of Own Assets -4 0 -6 2001 2003 2005 2007 2009201120132015 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC. Sources: FDIC and Flow of Funds (Haver Analytics). Note: Servicing refers to servicing of residential mortgages. Note: Merger-adjusted annual growth in 1–4 family residential mortgages “Serviced mortgages as percent of own assets” is computed only for banks outstanding at banks, and unadjusted growth of total U.S. 1–4 family servicing mortgages. Most community banks stopped reporting these items residential mortgages outstanding. starting in 2017 with the introduction of the new FFIEC Call Report 051.

Moreover, the percentage of community banks that service fees of any kind, including mortgage servicing fees, stood 1–4 family residential mortgages owned by others (a at 35 percent through 2019, slightly above its 2016 level. category that includes mortgages those banks originated and sold to a GSE with servicing retained) increased Noninterest Expense of Mortgage Specialists more or less steadily from 2001 to 2016, the last year Increased Relative to Other Banks After the most institutions reported these data, going from about Banking Crisis 11 percent of community banks in 2001 to about 26 percent The relatively robust continued participation of in 2016 (Chart 5.11).10 Data that were still being reported community banks in mortgage lending and servicing in 2019 provided no indication that community bank depicted in Charts 5.9–5.11 should not be taken to suggest mortgage servicing had dropped off after 2016. Specifically, that the mortgage and servicing rules had no effects on the percentage of community banks reporting servicing community banks. As noted above in Box 5.1, aggregate banking trends can mask developments affecting subsets 10 Most community banks began reporting using the FFIEC 051 Call of the industry. We consider—and find some evidence that Report form in 2017 . That form does not include the mortgage may be consistent with—two effects. One is the possibility servicing information depicted in Chart 5 11. .

5-12 FDIC Community Banking Study ■ December 2020 Chart 5.12

Community Bank Noninterest Expense by Degree of Mortgage Specialization Noninterest Expense to Assets Percent Community Banks With Mortgages as Percent of Assets of: 3.8 0 – 5 3.6 5 – 15 15 – 30 3.4 >30 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1996 1998 2000 2002 2004 2006 200820102012201420162018 Source: FDIC. Note: Full year noninterest expense as percent of average 5-quarter trailing assets; mortgages refers to 1–4 family residential mortgages. Gray bars denote recession periods.

Chart 5.13

Community Bank Noncurrent Loans by Degree of Mortgage Specialization Noncurrent Loans Plus ORE to Assets Percent 4.5 Banks With Mortgages as Percent of Assets of: 4.0 0 – 5 5 – 15 3.5 15 – 30 3.0 >30 2.5 2.0 1.5 1.0 0.5 0.0 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 Source: FDIC. Note: “Noncurrent” is 90 days or more past due or nonaccruing. “ORE” is other real estate owned. “Mortgages” is 1–4 family residential mortgages. Gray bars denote recession periods. that the new mortgage and servicing rules caused banks in associated increase in noninterest expense could be most the mortgage-lending business to incur greater expense readily observed for them.11 Chart 5.12 depicts noninterest for regulatory compliance. The other is the possibility that expense trends at community banks segmented into the desire to avoid such increased expense caused some four groups according to residential mortgage lending banks, particularly those with smaller mortgage programs, concentration relative to assets. Banks with residential to reduce or exit mortgage lending. mortgage concentrations greater than 30 percent of assets

It seems probable that banks with a substantial 11 The author is indebted to Nathan Hinton and Kevin Anderson, commitment to mortgage lending would be most likely whose internal FDIC research in 2016 analyzed noninterest expense of community bank mortgage specialists compared with community to stay in the business and absorb whatever additional banks having other degrees of residential mortgage concentration . compliance costs are necessary, and probable also that any Their research included preparing charts similar to Charts 5 12. and 5 13. in this chapter .

FDIC Community Banking Study ■ December 2020 5-13 are deemed mortgage specialists. The chart depicts an The analysis will shed only indirect light on the subject. inversion in noninterest expense ratios across the groups Call Reports of most small banks do not contain data on over time. The two highest mortgage concentration groups mortgage originations (Box 5.2 discusses the limitations— had the two lowest expense ratios pre-crisis, but post- for our purposes—of Home Mortgage Disclosure Act crisis they had the two highest expense ratios. Mortgage data on mortgage originations).12 Since mortgages can specialists had the lowest noninterest expense ratios stay on a bank’s balance sheet for many years, declines pre-crisis, and from 2014 to 2019 had the highest. in outstanding mortgage balances or mortgage interest income from one year to the next may mean the bank The post-crisis inversion of the relationship between exited the business, or may mean more mortgages paid noninterest expense ratios at mortgage specialists and off that year than were originated, or that mortgages were other banks is optically consistent with the hypothesis sold rather than held. Given these limitations, the analysis that mortgage-related compliance costs increased will view sustained annualized reductions in mortgage as a result of the post-crisis regulations, but other balances over a period of years as an imperfect proxy for a factors may have been as or more important. Mortgage strategic decision to scale back or exit mortgage lending. specialists may have been more likely to focus on building The analysis evaluates whether substantial annualized technological capabilities to compete with online and reductions in mortgage balances were more likely for mobile technologies pursued by others in this segment. banks that either were small in absolute size, or had small Also, as indicated in Chart 5.13, increases in noncurrent mortgage operations relative to their size. This approach is loans and other real estate during the crisis, while not intended to evaluate the idea that increases in regulatory as pronounced at banks with higher concentrations compliance costs may have made it less economical to in residential mortgages as they were at other banks, operate a small mortgage business. lingered longer. Higher levels of these troubled assets at banks in the two highest mortgage concentration Box 5.2 Home Mortgage Disclosure Act Data: groups for much of the post-crisis period may be part of Findings and Coverage Limitations the reason that the noninterest expense ratios of these two groups stayed higher than at other banks during the Unless banks are exempt under Regulation C, they must report originations of 1–4 family residential mortgages period 2013–2019. pursuant to the Home Mortgage Disclosure Act (HMDA). Research on mortgage trends based on HMDA An Unusually High Percentage of Small Mortgage data generally does not find aggregate reductions in Lenders Reduced Their Mortgage Holdings in the originations of purchase residential mortgage loans among reporting banks during the post-crisis period Years After the Banking Crisis (see, for example, Bhutta and Ringo (2016)). We next consider the possibility that some banks reduced Among the banks exempt from HMDA reporting, or exited the mortgage business to avoid regulatory however, are those that do not have a home office or compliance costs associated with the new rules. The branch in a metropolitan statistical area, and those results of banker surveys suggest this possibility. In one that originated fewer than 25 home purchase loans in survey (American Bankers Association (2016)), for example, either of the two preceding years. This exclusion of 33 percent of respondents in 2014, and 24 percent of small and rural mortgage lenders from reporting serves to limit the usefulness of HMDA data for purposes of respondents in 2015, stated that regulation was having an this chapter. “extreme negative impact” on their residential mortgage lending business. Other surveys and anecdotal reports For the current HMDA reporting criteria, see Federal stated that many community banks were considering Financial Institutions Examination Council, “A Guide to HMDA Reporting: Getting it Right,” at https://www. exiting mortgage lending altogether. ffiec.gov/hmda/pdf/2020guide.pdf.

The results shown in Charts 5.9–5.11 make clear that community banks, in the aggregate, have by no means 12 exited residential mortgage lending. Nevertheless, it is Call Report schedule RC-P requires reporting of mortgage originations by banks with assets exceeding $1 billion or banks possible that some community banks may have done so, that originated more than $10 million of mortgages in each of the and Call Report data will help us explore this possibility. two preceding quarters . Call Report form FFIEC 051, filed by most community banks, does not include schedule RC-P .

5-14 FDIC Community Banking Study ■ December 2020 Table 5.4 Changes in Mortgage Holdings of Community Banks, 1995–2019

Past Due and Share of Nonaccrual Loans Community Share of Average and Other Real Banks With Date Range Annualized Change in Number of Community Community Estate Owned as a Positive Growth (Year End of 1–4 Family Mortgage Community Banks Bank Assets Share of Assets in Other Loans Each Year) Loan Portfolio Banks (Percent) (Millions $) (Percent) (Percent) As of December 31, 2013 Increase 4,371 69 353 1 .43 91 -0 .1 Percent to -4 .9 Percent 938 15 253 2 .16 79 2013–2019 -5 .0 Percent to -9 .9 Percent 339 5 234 4 .13 63 Less Than -10 percent 362 6 223 2 .45 60 As of December 31, 2007 Increase 4,665 61 239 0 .86 70 -0 .1 Percent to -4 .9 Percent 1,514 20 250 1 .14 44 2007–2013 -5 .0 Percent to -9 .9 Percent 693 9 229 1 .25 38 Less Than -10 percent 444 6 246 2 .32 37 As of December 31, 2001 Increase 6,095 71 177 0 .66 92 -0 .1 Percent to -4 .9 Percent 1,277 15 143 0 .74 80 2001–2007 -5 .0 Percent to -9 .9 Percent 509 6 169 0 .85 70 Less Than -10 percent 436 5 211 1 .08 60 As of December 31, 1995 Increase 8,433 81 123 0 .88 93 -0 .1 Percent to -4 .9 Percent 830 8 114 1 .08 81 1995–2001 -5 .0 Percent to -9 .9 Percent 359 3 143 1 .35 80 Less Than -10 percent 311 3 125 1 .96 65 Source: FDIC . Note: Table does not include community banks that stopped reporting in 1996, 2002, 2008, or 2014 or that did not hold 1–4 family mortgages at year ends 1995, 2001, 2007 or 2013 . For such banks no annualized change in mortgage holdings could be calculated . Mortgage changes are annualized so that cumulative changes during the full date ranges would be larger .

Table 5.4 considers four six-year periods, and groups the similarities between the post-crisis period and earlier community banks existing at the beginning of each of periods. For example, in both the six post-crisis years the four periods according to their annualized percentage starting in 2013 and the six pre-crisis years starting change in residential mortgages during that period, or in 2001, about 26 percent of community banks had during their remaining existence, whichever was shorter. annualized reductions in mortgage holdings. In all four Thus, for example, a bank in the mortgage growth category of the six-year periods, community banks that reduced corresponding to annualized reductions of 10 percent mortgage holdings tended to have higher levels of or more would have reduced its mortgages by well over noncurrent loans and other real estate. In all periods 50 percent if it existed for all six years of a period, a except for the banking crisis, the majority of community reduction that quite possibly reflected its exit from the banks with annualized reductions in mortgages had business. Banks that reported no mortgages at the start annualized increases in their other loans. This suggests of a period, or that stopped reporting within one year of that the reasons for the reductions in mortgage loans the start of a period are not included, since no annualized may often have been specific to that business line rather change in mortgages could be computed for them. than to bank-wide or local economic issues. Examples of issues specific to mortgages in the post-crisis period

Table 5.4 indicates that with regard to community bank could include, for example, risks associated with holding reductions in mortgage holdings, there are a number of long-maturity assets on balance sheet in a low interest rate

FDIC Community Banking Study ■ December 2020 5-15 Chart 5.14

Percentage of Small Mortgage Lenders Materially Reducing Mortgage Holdings Banks With Annualized Reduction in Mortgages of 5 Percent or More Percent of Small Lender Group 60 1995 – 2001 2001 – 2007 2007 – 2013 50 2013 – 2019 40

30

20

10

0 Assets < $100 million Res Mtg < 5 percent of Assets Res Mtg < $1 million Source: FDIC. Note: Bars represent the percentage of community banks in each small lender group that reduced 1–4 family residential mortgage (Res Mtg) holdings at an annualized rate of at least 5 percent during the six-year period. Sustained reductions in balance-sheet amounts of residential mortgages are an imperfect proxy for reductions in residential mortgage originations but do not definitively establish a reduction in originations. environment. Similar to the overall picture suggested by lender groups that subsequently reduced their mortgage Charts 5.9 and 5.10, however, the overall picture suggested holdings at an annualized rate of 5 percent or more during by Table 5.4 does not support the idea that unusual the period. The proportion of small lenders substantially numbers of community banks in the aggregate were exiting reducing their mortgages increased with each successive mortgage lending during the post-crisis period. six-year period, and has been much higher during the post-crisis period even than during the 2008–2013 banking Table 5.4 does, however, suggest that during the post-crisis crisis. During the post-crisis period, moreover, while period, the size distribution of banks that were reducing about 11 percent of all community banks had annualized their mortgage holdings became skewed toward smaller reductions in mortgages of 5 percent or more (Table 5.4), banks. Specifically, during 2013–2019 the average asset over 30 percent of community banks with mortgages less size of community banks with annualized increases in than 5 percent of assets, and over 50 percent of community mortgage holdings was $353 million, exceeding by at least banks with mortgages less than $1 million, had annualized $100 million the average asset size of community banks reductions of this magnitude. that reduced their mortgage holdings. In the other three periods, in contrast, no systematic differences in asset size In short, during the post-crisis period small mortgage are evident in the table between banks that were reducing lenders had sustained material reductions in mortgage their mortgages and those that were increasing them. lending more frequently than larger community bank mortgage lenders did, and more frequently than small In fact, Chart 5.14 indicates, during the post-crisis period mortgage lenders had in previous periods. There may small mortgage lenders reduced their mortgage holdings be many reasons for a bank’s balance-sheet holdings with greater frequency than in any previous period. To of mortgages to exhibit a sustained decrease, including anticipate the discussion, the chart suggests that operating increased sales to the GSEs (as noted above, increased a small mortgage program or making mortgages as an sales to GSEs may themselves be driven by regulatory occasional customer accommodation may be becoming less considerations given the Qualified Mortgage safe harbor economical over time. The chart gives information about for such loans, or by a desire to avoid the interest-rate risk three possible definitions of a small mortgage lender, and associated with holding mortgages on the balance sheet). the patterns are robust to the definition used: community Nonetheless, the strong connection between reduced banks with assets less than $100 million, those with mortgage holdings and banks’ asset size and scope of mortgages less than 5 percent of assets, and those with mortgage operations suggests there may be factors at total mortgages outstanding less than $1 million. The chart work that are making it less economical for a bank to have reports the proportion of banks in each of these small- a small mortgage lending function. The factors that most

5-16 FDIC Community Banking Study ■ December 2020 Chart 5.15

Consumer Lending of Community and Noncommunity Banks Consumer Loans to Assets Percent Community Banks 14 Noncommunity Banks

12

10

8

6

4

2

0 1984 1988 1992 1996 20002004200820122016 Source: FDIC. Note: Consumer loans (which do not include 1–4 family residential mortgage loans) at year end as percent of assets, 1984–2019. Gray bars denote recession periods. readily suggest themselves are changes in financial and in 2019, down from about 30 percent in 2001. And even information technology (including increased competition for community banks that did report credit card loans, from nonbank entities) that promote commoditization of throughout the 2001–2019 period those balances totaled retail lending, and regulatory compliance costs resulting less than one-half of 1 percent of those banks’ assets. from the large volume of new mortgage rules. It is not possible to draw firm conclusions about the relative New disclosure and opt-in requirements regarding importance of these factors. overdraft programs are likely relevant to most community banks. Starting in 2015, institutions with assets of $1 billion Many Important New Rules Addressed or more that offer one or more consumer deposit account Consumer Credit and Retail Payments products have had to report overdraft charges on consumer accounts. The percentage of community banks in this size Another important group of rules implemented in the group reporting overdraft fees declined from 83 percent in 2008–2019 period addressed the broad category of 2015 to 77 percent in 2019, while the amount of such fees consumer credit and retail payments. Appendix B identifies (for banks reporting them) decreased modestly during and summarizes 27 distinct final rules in this category, the same period, dropping from 11 basis points of deposits rules that, broadly speaking, created rights and protections to 9 basis points of deposits.13 Downward pressure on for consumers, and obligations for lenders, related to credit service charges appears to be a long-term trend. From 2001 cards and other consumer credit, the use of credit reports, through 2019, deposit service charges at community banks customer overdrafts, gift and prepaid cards, remittances, decreased from 38 basis points of deposits to 19 basis points and retail foreign exchange. of deposits; the corresponding decrease at noncommunity banks was from 67 basis points to 28 basis points. Although consumer loans constituted less than 3 percent of community bank assets throughout the post-crisis period International remittance transfers, which historically (Chart 5.15), almost all community banks have at least had been exempt from federal consumer protection some consumer loans and need to be aware of changes to laws, became subject to a disclosure and consumer consumer regulations.

13 The overdraft fees reported by this category of institutions are Requirements specific to credit card lending applied reported on Call Report schedule RI, memorandum item 15 . a), to a relatively small set of community banks. About “Consumer overdraft-related service charges levied on those transaction account and nontransaction savings account deposit 16 percent of community banks reported credit card loans products intended primarily for individuals for personal, household, or family use ”.

FDIC Community Banking Study ■ December 2020 5-17 Chart 5.16

Community Banks Meeting Asset Threshold for 18-Month Examination Cycle Percent of Community Banks Eligible Based on Asset Size 100 90 80 70 60 50 40 30 20 10 0 1991 1993 1995 1997 1999 2001 2003 20052007200920112013201520172019 Source: FDIC. Note: Some banks qualifying for an 18-month examination cycle based on their asset size may be examined more o­en based on supervisory considerations. The asset size threshold at year end 2019 was $3 billion. Gray bars denote recession periods.

protection regime, although institutions making cycle) was increased from $500 million to $1 billion fewer than 100 remittances per year were exempt from and later to $3 billion. As of year-end 2019, more than these requirements. At mid-2019, about 10.5 percent of 98 percent of community banks met the asset size community banks reported providing more than 100 threshold for an 18-month examination cycle (Chart 5.16). international remittances per year, up slightly from In 2019, the loan size threshold above which federally 9 percent at mid-2014.14 related mortgage loans require an appraisal was increased from $250,000 to $400,000 for residential mortgages and Numerous Other Regulations Were Finalized from $250,000 to $500,000 for commercial mortgages. During the Years 2008–2019 Other important safety-and-soundness rule changes This brief overview of the remaining rules listed in affected community banks to varying degrees. Risk Appendix B may be taken as a reminder that there were retention rules, which require securitizers to retain a many important rule changes during 2008–2019 with 5 percent loss exposure to assets they securitize unless which community banks had to be familiar. one of numerous exceptions applies, likely directly affect few community banks, but those interested in becoming The Federal Banking Agencies Implemented active securitizers would need to be knowledgeable about Important Changes to Safety-and–Soundness Regulations these rules. The Volcker Rule’s statutory prohibition on proprietary trading and ownership or sponsorship of There were a number of regulatory changes to safety- hedge funds or private equity funds was finalized in 2013, and-soundness rules affecting small banks during the and in 2018 it was statutorily rescinded for most banks period 2008–2019, many of them statutory. Derivatives with assets below $10 billion. Similarly, company-run exposures were incorporated into the national bank legal stress testing requirements for banks with assets greater lending limit; regulations governing banks’ permissible than $10 billion were implemented in 2012, but the asset investments were de-linked from credit ratings; a threshold was statutorily raised in 2018. Very large portion of reciprocal deposits was excluded under certain community banks or those considering acquisitions that circumstances from being defined as brokered deposits; would cause them to exceed $10 billion in assets would and the maximum asset threshold for eligibility for an have needed to comply with or consider these stress- 18-month examination cycle (rather than a 12-month testing requirements.

14 These data are reported only on the June 30 Call Report .

5-18 FDIC Community Banking Study ■ December 2020 Chart 5.17

Derivatives Activity of Community Banks Percent of Banks and Assets 25 Percent of Community Banks With Derivatives 20

15

Gross Notional, Percent of 10 Assets for Community Banks With Derivatives

5

0 1995 1997 1999 2001 2003 2005 2007200920112013201520172019

Source: FDIC. Note: Gross notional amounts of derivatives held for trading and non-trading by community banks with positive reported derivatives holdings. Gray bars denote recession periods.

Bank Secrecy Act and Law Enforcement for institutions to disclose to their deposit sweep Responsibilities Increased customers how their sweeps would be treated by the FDIC in the event of the bank’s failure. Another 2008 Banks have responsibilities to take actions and provide rule, amended in 2017, requires that banks in a troubled information in support of law enforcement, and three rules condition, upon written notice from the FDIC, be able to put in place since 2008 increased these responsibilities. provide specified information regarding their Qualified One was a requirement that U.S. financial firms that Financial Contracts (or QFCs, which include swaps, participate in designated payment systems (a group that securities financing transactions, and repurchase includes most banks) establish and implement policies agreements) to the FDIC on request as of the end of a and procedures that are reasonably designed to prevent business day. The QFC rule does not appear to have had any payments to gambling businesses in connection with ancillary effect of dampening community banks’ use of unlawful internet gambling. The second established derivatives: on the contrary, the proportion of community specific suspicious activity reporting and information banks that hold derivatives increased fairly steadily from collection requirements on providers of prepaid access about the year 2000 through 2019 (Chart 5.17). devices such as cards, although the requirements generally exempted small balance products (balances below $1000). The Dodd-Frank Act Made Two Important Changes to The third was the customer due diligence rule, which the Pricing of Bank Products and Services requires financial institutions to identify and verify the identity of the beneficial owners of companies opening The mortgage and consumer credit rules described above accounts, understand the nature and purpose of customer contain a number of fee limits or regulatory requirements relationships in order to develop customer risk profiles, that are triggered by levels of fees or interest rates. Two and conduct ongoing monitoring to identify and report other notable rules from the 2008–2019 period dealt suspicious transactions and, on a risk basis, to maintain with the pricing of bank products or services. In 2011 the and update customer information. Federal Reserve implemented the Dodd-Frank Act’s limits on the interchange fees of banks with assets greater than Some Rules Were Related to the FDIC’s $10 billion, an asset size group that has included some Responsibilities for Resolving Failed Banks community banks. Also in 2011, the Dodd-Frank Act’s repeal of the statutory prohibition against banks’ paying Some rules were driven by the FDIC’s resolution interest on demand deposits took effect. responsibilities. A 2008 rule introduced the requirement

FDIC Community Banking Study ■ December 2020 5-19 Some Rules Affect Bank Competition and and community development matters. Among the Industry Structure more significant of these were two rules that together implemented the flood insurance provisions of the Some rules reflect statutory goals for the avoidance Biggert-Waters Act, which among other things clarified of undue concentration or anti-competitive practices when banks could and should accept private flood in banking. One such rule from the 2008–2019 period insurance policies. Several rules during the 2008–2019 implemented the Dodd-Frank Act’s prohibition on period addressed back-office functions, including issues acquisitions if the resulting company would have more arising from the banking system’s ongoing migration from than 10 percent of all U.S. financial institution liabilities. paper-based to electronic payments. These included rules Another rule from 2019 eased restrictions on management dealing with paper and electronic check processing and interlocks by permitting a management official to serve dispute resolution, funds availability, the settlement cycle at two unaffiliated banks unless both have more than for securities transactions, and other matters. $10 billion in assets, or unless both operate in the same geographic area. Community Bank Exit and Entry May Significant Requirements Took Effect Regarding Have Been Affected by the Pace of Financial Reporting and Auditing Regulatory Change This analysis of regulatory changes has focused on A significant development during the 2008–2019 period individual rules and individual balance-sheet and income- was a 2009 FDIC rule applicable to insured institutions statement categories, thus far without consideration of the with assets exceeding certain thresholds. Consistent possible totality of effects. Trends in bank exit and entry with the Sarbanes-Oxley Act, the rule, among other may shed light on such total or cumulative effects. Rates things: (1) requires disclosure of an institution’s of exit from the banking industry, and entry into it, can be internal control framework and material weaknesses; viewed as high-level indicators of how bankers view the (2) requires management’s assessment of compliance economic prospects of banking franchises given a wide with laws and regulations; (3) clarifies the independence range of factors, including regulatory changes. standards applicable to accountants; (4) establishes a variety of requirements regarding audit committees; and A previous section of this chapter showed that smaller (5) establishes criteria for institutions to comply with the community banks have had higher proportionate requirements at a holding company level. For a holding noninterest expense than larger community banks and company’s insured subsidiaries to be able to satisfy the that any given increment of overhead expense would audit requirements at the holding company level, the assets weigh more heavily on their bottom lines. Accordingly, it of the subsidiaries must be at least 75 percent of the holding is not unreasonable to think that changes in regulatory company’s consolidated assets. Institutions covered by the requirements that involve a significant learning curve, rule are generally those with at least $1 billion in assets legal or consulting fees, or additional staff time could for purposes of internal control assessments and at least tend to depress small-bank profitability relative to other $500 million for purposes of other requirements. banks, with the indirect result of encouraging some small banks to exit the banking industry, or of discouraging the More recently, in 2019, the federal banking agencies chartering of new small banks. expanded the eligibility of institutions that could file the most streamlined version of the Call Report, the FFIEC As Chapter 2 notes, banking consolidation has been 051, to include insured depository institutions with total underway since the 1980s, with the most rapid rate of assets of less than $5 billion that do not engage in certain consolidation occurring in the late 1990s. But whereas the complex or international activities. consolidation of the 1990s had been driven by the ongoing relaxation of branching restrictions, a relaxation that Other Regulations Addressed Flood Insurance, Back-Office Functions, and Other Matters resulted in consolidation of many multi-bank holding companies under a smaller number of charters, a new Appendix B documents 13 other rules (or in a few cases and important factor in the decline in the number of interagency questions and answers) from the 2008–2019 institutions since the 2008–2013 banking crisis was the period dealing with assorted other consumer protection relative dearth of new charters. Chapter 2 also notes that

5-20 FDIC Community Banking Study ■ December 2020 Chart 5.18

Community and Noncommunity Bank Exit From Banking by Una iliated Merger or Self-Liquidation Acquired by Unailiated Bank or Self-Liquidating Percent of Existing Institutions 12 Percent Exiting: Community Banks 10 Noncommunity Banks 8

6

4

2

0 1984 1988 1992 1996 20002004200820122016 Source: FDIC. Note: Percent of all banks existing at year end that were acquired by an unailiated institution or self-liquidated within the next year. Gray bars denote recession periods.

Chart 5.19

Exit Rates of Metro and Non-Metro Community Banks Acquired by Una iliated Bank or Self-Liquidating Percent Exiting: Percent of Existing Institutions Metro Community Banks 6 Non-Metro Community Banks

5

4

3

2

1

0 1984 1988 1992 1996 2000 2004 2008 2012 2016 Source: FDIC. Note: Percent of all banks existing at year end that were acquired by an una iliated institution or self-liquidated within the next year. Gray bars denote recession periods.

consolidation is not purely a community-bank trend—for banks as reflected in their initial equity. Chart 5.18 depicts in fact noncommunity banks have consolidated at faster the annual percentages of community banks exiting rates than community banks—and in addition that when the banking industry, either through acquisition by an community banks have been acquired, the acquirers have unaffiliated institution or by self-liquidation. This type of mostly been other community banks. exit would seem to reflect a decision by bank ownership that the bank’s continued existence as an independent The two post-crisis developments with which this entity was no longer financially advantageous. The chart section of the present chapter is concerned are the shows that community banks were exiting the banking historically high proportion of community banks exiting industry at the fastest rates since 1984 (although, as the the banking industry in the years 2014–2019, and an chart also makes clear, not as fast as exit rates sometimes apparent increase in the target asset size of new small observed for noncommunity banks), with an average exit

FDIC Community Banking Study ■ December 2020 5-21 Chart 5.20

Community Bank Exit Rates by Age of Bank Acquired by Una iliated Bank or Self-Liquidating Percent of Existing Institutions Percent Exiting: Age < 10 years 16 Age 10 – 40 years Age > 40 years 14

12

10

8

6

4

2

0 1984 1988 1992 1996 20002004200820122016 Source: FDIC. Note: Percent of all banks existing at year end that were acquired by an una iliated institution or self-liquidated within the next year. Gray bars denote recession periods.

Chart 5.21

Exit Rates of Very Small Community Banks Acquired by Una iliated Bank or Self-Liquidating Percent Exiting: Percent of Existing Institutions Community Banks With 8 Assets Less Than $50 Million All Other Community 7 Banks 6

5

4

3

2

1

0 1984 1988 1992 1996 20002004200820122016 Source: FDIC. Note: Percent of all banks existing at year end that were acquired by an una iliated institution or self-liquidated within the next year. Gray bars denote recession period.

rate between 2014 and 2019 of just over 4 percent, compared Regulatory factors also have been asserted to affect entry with the previous high of 3.7 percent in 1997. Post-crisis exit into the banking industry. Many commentators have rates were particularly high for metro banks (Chart 5.19); stated that the decline in the number of new charters for young banks (Chart 5.20—for this chart, young banks after the 2008–2013 banking crisis was caused partly by are defined as those less than ten years old; accordingly, the regulatory environment, while other commentators they were chartered shortly before or during the 2008–2013 have emphasized economic factors. Adams and Gramlich banking crisis); and for the very smallest banks (Chart 5.21). (2014), for example, contains an analysis of economic These charts also make clear, however, that for community factors underlying chartering activity. Rather than banks that are rural, older, or larger in size, post-crisis exit re-examining an issue that has been studied at length rates also were at or near historic highs. elsewhere, we consider how market perceptions have

5-22 FDIC Community Banking Study ■ December 2020 Chart 5.22

Initial Equity of New Small Banks Mean Initial Equity of De Novo Small Banks $ Millions 40 35 30 25 20 15 10 5 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC. Note: Equity capital reported as of banks’ first Call Report filing, analysis limited to de novo banks with initially reported equity capital less than $100 million. There were no de novo banks in 2011, 2012, 2014 and 2016. Three new bank charters established in 2011 in connection with bank failures are not included in this chart as de novo banks.

Chart 5.23

Percentage of New Banks With Initial Equity Capital Less Than $20 Million De Novo Banks With Initial Equity Less Than $20 Million Percent of all De Novos 100 90 80 70 60 50 40 30 20 10 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Source: FDIC. Note: There were no de novo banks in 2011, 2012, 2014, and 2016. In 2015 and 2017 there were de novo banks but none with initial equity less than $20 million. Three new bank charters established in 2011 in connection with bank failures are not included in this chart as de novo banks.

changed as reflected by initial investment in new banks. and discontinuously, from $11.8 million during the period Chart 5.22 provides indirect evidence that the target asset 2000–2010 to $29.6 million starting in 2015. size of new small banks increased during the post-crisis period. The chart displays the mean equity reported in the The marked rise in initial equity is not attributable to first quarterly financial reports of new banks, as a proxy inflation, which was muted, or to changes in regulatory for the owners’ initial capital investment. New banks with capital requirements, for those requirements had not initial equity of more than $100 million are not included, changed enough to explain a change in initial capital of since their relative infrequency and large size would mask this magnitude. Inasmuch as initial equity is intended to patterns of interest for smaller banks. Mean initial equity allow the bank to achieve and support its planned asset for these smaller banks increased, somewhat abruptly size, it appears reasonable to suppose that the relatively few new banks chartered after the crisis had higher

FDIC Community Banking Study ■ December 2020 5-23 projected asset sizes than new banks chartered before the in those industries may tend to favor larger banks. crisis. The trends in Chart 5.22 suggest that proponents Keeping pace with new technologies also may be easier of new banks believed that the scale of operations needed for larger banks. Challenges in arranging for appropriate to make a new bank successful had increased in the post- management succession, sometimes in situations crisis period. An increase in the target size of new banks, involving the generational transfer of family-owned in turn, could plausibly be associated with scale economies, banks, in which the following generation is not interested attributable at least partly to regulatory compliance costs.15 in taking on the operation of the family’s bank, have been cited by some bankers as a factor that may influence some The preceding discussion should not be taken to imply banks to seek an acquirer. Commoditization of retail that new small banks can no longer be chartered or cannot lending also likely favors larger financial firms whose be successful. As indicated in Chart 5.23, in 2018 and 2019 average cost structures are lower and that deploy new some banks were chartered with initial equity of less than technology. $20 million, a level of initial equity that had characterized the overwhelming majority of new banks chartered in the A shared characteristic of some of the important factors years 2000–2009. driving developments in banking—changes in customer demographics, in the nature of marketplace competition, Summary in technology, and in regulation—is that all are factors external to a bank that can cause the bank to have to Bankers have sometimes characterized the regulatory change the way it does business. All may involve a need costs they incur as being difficult to attribute to any one for evolving capabilities, consultants, or other specialized set of rules, but as the cumulative effect of many rules. staff, and all may involve relatively higher fixed costs or The review in this chapter and its appendix of a partial generally greater challenges for smaller institutions. Such list of regulatory actions taken by six federal agencies factors evolve continually, making it hard from financial (often implementing statutory mandates from Congress) data alone to know whether—and in what degree—to from 2008 through 2019 makes clear that merely keeping attribute any particular trend to changes in regulation, or current on banks’ regulatory requirements as they to one or more of the other factors. evolve cumulatively through time is a daunting task for anyone, and certainly for a small bank with modest staff Finally, it is important to emphasize that this study and resources. views regulations only through the lens of their effects on community banks; a discussion of the policy goals Regulatory compliance costs may be one of a number of Congress has sought to achieve with its statutes, or how factors contributing, for example, to higher rates of exit well the regulations have achieved those goals, is beyond from the banking industry by community banks; to an the scope of the analysis. Observations in this study about apparent increase in the target asset size of new small the effects of rules on community banks should thus not be banks; or to a pronounced increase in the proportion of taken as criticisms of those rules. The overall thrust of the small residential mortgage lenders that are reducing their analysis, however, does support the idea that if the societal residential mortgage holdings. Most likely other factors benefits of a thriving community banking sector are to be are also very important contributors to these trends, and preserved, it is important that regulations achieve their we draw no conclusions about the importance of any of public policy goals in ways that accommodate, to the extent these other factors compared with changes in regulatory appropriate, the business models and learning curves of compliance costs. Business consolidation is occurring in smaller institutions with limited compliance resources. many industries, not just banking, and larger companies

15 See Jacewitz, Kravitz, and Shoukry (2020) for a recent analysis of bank scale economies .

5-24 FDIC Community Banking Study ■ December 2020 Box 5.3 Regulatory Developments During the COVID-19 Pandemic

A brisk pace of regulatory activity has continued during the pandemic, with a focus on rules and programs that encourage and facilitate banks’ provision of financial services to their customers. An important statutory backdrop for some of the pandemic-related rules was the Coronavirus Aid, Relief, and Economic Security (CARES) Act, a $2.2 trillion economic relief package signed into law on March 27, 2020. Examples of pandemic-related rules and federal programs affecting community banks include:

• Establishing the Federal Reserve’s Paycheck Protection Program Liquidity Facility to provide liquidity to banks to support their participation in the PPP; • Extending the regulatory capital transition period for banks adopting the Current Expected Credit Loss Accounting Standard; • Temporarily reducing the Community Bank Leverage Ratio threshold to 8 percent as required by the CARES Act; • Modifying capital rules to neutralize the regulatory capital effects of banks’ participating in the PPP, and establishing a zero-percent risk weight for those loans as required by the CARES Act; • Deferring certain required real-estate appraisals and evaluations for up to 120 days after loan closing; and • Modifying FDIC deposit insurance premiums to mitigate the effects of banks’ participating in the PPP.

FDIC Community Banking Study ■ December 2020 5-25

Chapter 6: Technology in Community Banks

From mobile banking to online lending, financial community banks and those with higher revenues to assets technology is reshaping how customers want to bank were most likely to have adopted certain technologies. and how banks can deliver products and services. For community banks at the forefront of this movement, the Future research into technology adoption will broaden latest technology-enabled products and services have our understanding of the key drivers, barriers, and risks become a necessity rather than a luxury. Other banks, associated with financial technology and its likely effect meanwhile, have charted a more conservative course, on the continuing success of community banking. adopting new technology only after it has settled into mainstream banking. Somewhere in between the early CSBS Survey Data Offer a Representative and late adopters lie the thousands of community banks Look at Community Banks that operate under different business models in different Since 2015, the Conference of State Bank Supervisors (CSBS) environments throughout the United States. has conducted an annual national survey of community banks to quantify underlying trends and issues of This chapter differentiates community banks on the basis importance. The 2019 survey, conducted between April and of their technology offerings, thereby contributing to a June of that year, generated responses from 519 institutions better understanding of the factors that influence, and that met the definition of “community bank” using Call are influenced by, banks’ decisions to adopt technology. Report data as of March 31, 2019.1 Existing research in combination with responses to the Conference of State Bank Supervisors 2019 National The banks examined in this chapter generally reflected Survey of Community Banks reveals several factors that the overall population of community banks at the time of were related with the adoption of technology. Among the survey. Respondents were spread across the country, these factors were a bank’s characteristics, its economic with branches in 43 states and the District of Columbia and competitive environment, and the attitudes and (Map 6.1). Table 6.1 shows that in distribution by size expectations of its leadership. In particular, larger (as measured by total assets and number of branches), Map 6.1

CSBS Survey Respondents by Location, First Quarter 2019

Census Region Number Percent West 68 13.1 Midwest 232 44.7 Northeast 56 10.8 Headquarters South 163 31.4 Branch

Sources: FDIC and Conference of State Bank Supervisors. Note: No respondents had their headquarters or a branch in Alaska.

1 For its survey, the CSBS defined “community bank” as an institution with less than $10 billion in total assets . Differences between that definition and the definition used in this study resulted in the exclusion of 52 institutions from the findings discussed in this chapter, relative to the summary of the survey results published by the CSBS .

FDIC Community Banking Study ■ December 2020 6-1 Table 6.1 Comparison of Surveyed Banks and by the survey included three that help banks in lending All Community Banks by Asset Size and Number of Branches, (online loan applications, online loan closure, and First Quarter 2019 automated loan underwriting) and four that provide Community Banks additional functionality to deposit accounts among other (Percent of Total) In Survey All functions (remote deposit capture, interactive teller Total Assets machines (ITMs), electronic bill payment, and mobile Less Than $100 Million 18 3. 25 .1 banking). A general description of each technology is $100 Million to $200 Million 24 9. 23 .9 included in Box 6.1. Several of the technologies date from $200 Million to $500 Million 30 1. 29 .2 the early to middle 2000s,3 whereas others, such as online $500 Million to $1 Billion 14 1. 12 .6 loan closure and ITMs, emerged in the middle to late 2010s. More Than $1 Billion 12 7. 9 .3 Overall, among community banks participating in the 2019 CSBS survey, adoption rates ranged from 4.8 percent Number of Branches for online loan closure to 90.9 percent for mobile banking 1 Branch 12 1. 19 .1 (Chart 6.1). 4 2 to 4 Branches 38 5. 39 .7 5 to 9 Branches 28 5. 24 .6 In addition to looking at whether community banks 10 to 19 Branches 14 8. 11 .7 offered a technology-enabled product or service, the 20 to 49 Branches 5 .2 4 .3 chapter combines the seven technology offerings into one 50 or More Branches 0 .8 0 .7 general technology-adoption measure. Specifically, the Sources: FDIC and Conference of State Bank Supervisors . measure categorizes each community bank as a “low,” “medium,” or “high” adopting bank on the basis of the the surveyed community banks generally reflected the number and type of technology products and services (out distribution of all community banks. As of March 31, 2019, of the seven included in the survey) that the bank offered on average, community banks in the CSBS survey held at the time of the survey. Products and services that were about $36 million more in assets than all community less common (those with an adoption rate of less than banks, and operated one more branch than all community 50 percent) received greater weight than those that were banks; these differences, however, were comparatively more common (those with an adoption rate of greater than small—4 percent and 12 percent of a standard deviation, 50 percent) so that banks that were “early adopters” of respectively.2 one or more less common technologies were more likely to be defined as high-adopting banks. For more detail, Adoption of Certain Technologies Varied see Box 6.1. Among Community Banks

Technology has a long history in banking, yet the data necessary to quantify its adoption and use are hard to obtain, particularly data on community banks. On the Call Report, banks do not report their use of, or spending for, technology, and information available through other 3 For example, Wells Fargo was one of the first U .S . banks to introduce regulatory filings is often not comparable across entities or mobile banking in 2001, although the bank discontinued the service is not required of many smaller institutions. shortly thereafter . Other large banks, including Citibank, Bank of America, and Wachovia, added similar services beginning in 2006 and 2007 . See Hamilton (2007) . First Tennessee Bank in Memphis was one Therefore, this chapter relies on responses to the CSBS of the first financial institutions to implement remote deposit capture survey that indicated whether a community bank offered in 2003 as a way to expand its deposit base . The Check Clearing for the 21st Century Act, which took effect in 2004, paved the way for the specific technology-enabled products or services at the further development of remote deposit capture by allowing image- time of the survey. The products and services covered based “substitute checks” to serve as the legal equivalent of an original check . See FDIC, “Remote Deposit Capture: A Primer” (2009) . 4 In addition to stating whether their bank offered a specific technology-enabled product or service, survey respondents indicated 2 Unless otherwise specified, this chapter uses Call Report data from whether they planned to offer, or to exit or substantially limit, the March 31, 2019, the quarter immediately preceding the collection of product or service within the next 12 months . For purposes of this survey data . This contrasts with other chapters of this study, which chapter, adoption status includes only a bank’s reported offering at generally use data through year-end 2019 . the time of the survey and does not account for intentions .

6-2 FDIC Community Banking Study ■ December 2020 Chart 6.1

Adoption Rate of Surveyed Community Banks by Product and Service, 2019

Mobile Banking 90.9

Electronic Bill Payment 83.2

Remote Deposit Capture 78.8

Online Loan Applications 37.2

Automated Loan Underwriting 11.6

Interactive Teller Machines 10.8

Online Loan Closure 4.8

0102030405060708090100 Percent of Adopting Community Banks Source: Conference of State Bank Supervisors.

Box 6.1. Process for Creating a General Technology-Adoption Measure

The CSBS survey asked community banks to state their intentions toward the seven technology-enabled products and services listed below.a

Automated loan underwriting – platform that retrieves and processes borrower data through a computer- programmed underwriting system to arrive at a logic-based loan decision

Electronic bill payment – ability for customers to transfer funds from a transaction or credit-card account to a creditor, vendor, or individual

Interactive teller machines – machines that provide customers with direct, real-time access to a bank teller via a video link

Mobile banking – service that allows customers to access account information and conduct transactions with their institution remotely via a mobile device (cell phone, tablet, etc.)

Online loan applications – portal for potential borrowers to electronically share items, such as identifying information, income, and bank account data, needed to process a loan application

Online loan closing – ability to electronically sign and complete documentation necessary to finalize a loan (note that some states do not allow full remote online notarization)

Remote deposit capture – service that allows a customer to remotely scan checks and transmit the images or data to a bank for posting and clearing

Each of these technology products and services were categorized as either “more common” (if offered by more than half of community bank respondents to the CSBS survey as shown in Chart 6.1) or “less common” (if offered by fewer than half of community bank respondents).

Table 6.1.1 includes a 5x4 matrix that depicts the number of community banks that offered different combinations of “more common” and “less common” technologies. For example, the cell in the first numbered column and row indicates that 14 community banks in the survey offered none of the “less common” or “more common” technologies; whereas, the last numbered column and row indicates that two community banks offered all four of the “less common” technologies and all three of the “more common” technologies. continued on page 6-4

a The definitions included in this chapter are for informational purposes . Community banks participating in the CSBS survey used their own interpretations when indicating whether the bank offered a product or service .

FDIC Community Banking Study ■ December 2020 6-3 Box 6.1, continued from page 6-3

Next, each cell and its corresponding banks were labelled “low-adopting” (tan-shaded cells in Table 6.1.1), “medium- adopting” (dark gold-shaded cells),” or “high-adopting” (dark blue-shaded cells.) The labels were chosen in a manner that divided banks evenly among the categories, to the extent possible, to allow for more equal comparisons across groups. Labels were also chosen so that “high-adopting” banks were more likely to offer a greater number of technologies and be early adopters of “less common” products and services. The result of the process by which the low-, medium-, and high-adopting schema was arrived at is depicted in the right-hand table of Table 6.1.1.

Table 6.1.1 Number of Technologies Offered by Adoption Category, 2019

“More “Less Common” Technologies Common” Technologies 0 1 2 3 4 Total Number % 0 14 2 1 0 0 17 Low-Adopting Banks 131 25 .2 1 32 5 1 1 0 39 Medium-Adopting Banks 193 37 .2 2 78 28 8 0 1 115 High-Adopting Banks 195 37 .6 3 154 132 48 12 2 348 Total 278 167 58 13 3 519 Sources: FDIC and Conference of State Bank Supervisors .

Research and Survey Responses Link extent that those aspects reflect differences in customer Several Factors With Technology Adoption demand for technology.

Existing research has identified several characteristics In addition to previous research, the CSBS survey data that differentiated banks that adopted earlier technologies offer another perspective on factors that may be relevant from those that did not. Studies of the ATM and internet to the adoption of financial technology. When asked banking, for example, found that larger banks adopted to describe the “most promising opportunities facing the technologies at a faster pace. Internet adoption was your bank regarding new technology,” community also associated with improved profitability, higher deposit banks focused more on their customers than on other service charges, increased use of certain deposits, and potential benefits, such as cost savings and efficiency higher average employee wages.5 Research on general gains. Phrases referencing customer-facing services, technology expenditures found that increased spending such as “mobile bank,” “remote deposit,” and “online in previous years led in later years to greater output— account,” were among those most often used by survey as measured by loans, deposits, and number of branches— respondents (Chart 6.2). The frequent appearance of the as well as to higher bank employment, even after bank size phrases “account open,” “new customer,” and “younger was accounted for.6 generation” further suggest that some community banks saw customer opportunities that extended beyond the Research also identifies several external factors linked banks’ existing base, and these banks might have been to technology adoption. Competition, as measured by the motivated by the potential for growth. The opportunities adoption decisions of nearby competitors, appeared to cited by community banks did not differ significantly influence banks’ decisions to adopt the ATM and mobile among low-, medium-, and high-adopters, as defined banking applications.7 Studies documenting a “digital above. divide” between age groups and between urban and rural areas suggest that certain aspects of a bank’s environment As Chart 6.2 also shows, community banks frequently may also play a role in the bank’s adoption decisions to the referenced cost, as well as the phrase “keep up,” to describe the “most difficult challenges” presented by new 5 Hannan and McDowell (1984); DeYoung, Lang, and Nolle (2007); Sullivan and Wang (2013); Dahl, Meyer, and Wiggins (2017) . technology. In some cases, banks used the phrase “keep 6 Feng and Wu (2019) . up” in the context of “keeping up” with competitors— 7 Dos Santos and Peffers (1998); He (2015) .

6-4 FDIC Community Banking Study ■ December 2020 Chart 6.2 Most Common Phrases Characterizing Technology’s Opportunities and Challenges What are the most promising opportunities facing What are the most diicult challenges facing your bank regarding new technology? your bank regarding new technology?

Mobile Bank 26 Keep Up 28 Account Open 23 Core Provider 18 Online Account 10 Cost Technology 14 Remote Deposit 9 Core Process 14 Loan Application 8 Cyber Security 12 New Customer 7 Cost Implement 12 Online Bank 7 Big Bank 7 Large Bank 7 Third Party 7 Younger Generation 7 Pace Change 6 Peer to Peer Payment 7 Core System 6

Sources: FDIC and Conference of State Bank Supervisors. Notes: Counts are based on the number of institutions that used a given phrase; 417 community banks responded with at least one opportunity; 440 community banks responded with at least one challenge; response rates did not vary significantly between low-, medium-, and high-adopting banks.

often larger banks. In other cases, community banks mentioned in just under half (48 percent) of the responses referred to the challenge of “keeping up” with the rapid provided by low-adopting banks, but also by about pace of technology development. 40 percent of high adopters.

References to cost were linked with technology, in general, Unlike their descriptions of opportunities and apart from as well as with the implementation of technology. In costs, responses from low- and high-adopting community addition to appearing in the most-common phrases, banks differed with respect to the challenges presented cost was also the single word most frequently used by all by technology. As Figure 6.1 also shows, low-adopting community banks to describe challenges (Figure 6.1). Use banks more frequently used words such as “security,” of the word “cost” was highest among low technology “regulation,” “risk,” and “compliance,” relative to high- adopters, but not by a significant margin: the word was adopting banks. High-adopting banks, on the other hand,

Figure 6.1 Most Common Single Words Describing Hardest Challenges of New Technology

Low-Adopting Banks High-Adopting Banks

compete security compete provider keep keep implement implement change resource pace staff service small cyber limit staff service risk employee expense costregulaon product system cost processor regulaon me core customer vendor expense compliance customer security

Sources: FDIC and Conference of State Bank Supervisors. Notes: Sized on the basis of the percentage of institutions using the word at least once in their response; 440 community banks responded with at least one challenge; response rates did not vary significantly by adoption category.

FDIC Community Banking Study ■ December 2020 6-5 more often used words such as “core,” “provider,” and “vendor,” which are associated with third-party service “Small bank with a small number of customers providers and, particularly, with core service providers. makes it difficult to justify the cost of new products.” —(Low-adopting) community-bank president Across all community banks, 46 (mostly medium- and high-adopting banks) cited their core systems or core “The cost of technology is prohibitive as well as service providers when describing the most difficult the implementation and training of staff to utilize challenges of new technology. Specifically, when referring technology to its full potential.” to their core systems, community banks noted limited —(High-adopting) community-bank executive access to desired products and services, integration with “Vendors move to[o] slow and for smaller banks we current systems, a lack of alternative providers, and speed are pushed to back of line.” to implementation. —(Medium-adopting) community-bank president

To further explore how banks that adopted technology Larger banks may also benefit from greater bargaining differed from those that did not, the chapter now power when purchasing technology. For example, a examines the links between technology adoption and technology service provider may be more willing to factors identified above: a bank’s size and revenues; customize a product or service for a larger institution the relationship between adoption and loans, deposits, because of the additional income and exposure the growth, and performance; the role played by a bank’s provider would receive, while offering little to no flexibility environment; and the role played by leadership’s attitudes to a smaller institution. and expectations.

On the other hand, bank size may have less of an effect Community Banks With More Assets and on technology adoption if the cost of adopting a certain Revenues Were Greater Technology Adopters technology has declined over time. This decline may be due to recent technologies’ need for less hardware or to the Existing research on the adoption of earlier technologies, possibility of obtaining cheaper or more widely available as well as the large number of survey responses that technology through service agreements with third parties. mentioned cost, suggest that a bank’s size and resources For further discussion of how banks obtain technology, were major determinants of its decision to adopt or not see Box 6.2. adopt different technologies.

Bank Size Was the Strongest Indicator of On average, high-adopting community banks in the CSBS Technology Adoption survey were larger than low- and medium-adopting banks. The average high adopter reported assets that Size may be associated with the adoption of technology if were $324 million greater than medium adopters and larger banks benefit from economies of scale by spreading $535 million greater than low adopters.9 (Differences in the fixed costs of adopting technology over a wider customer the median were smaller, but still large, with the median base. Banks also tend to hire more employees as they grow high-adopting bank holding $228 million more assets than in size, making it easier for some workers to specialize the median medium-adopting bank and $344 million more in technology-specific functions, such as development assets than the median low-adopting bank.) and maintenance, vendor research and selection, risk management, and compliance. Although many people Differences in technology adoption were most evident associate economies of scale with large regional and between the largest and smallest community banks. national banks, other work cited by this study found that Only 6 percent of community banks with assets less than community banks generally realize most of the benefits of $100 million were high adopters, compared with 70 percent scale by the time they reach $600 million in assets.8 This makes it plausible to suggest that economies of scale do not 9 For some bank factors described in the chapter, including asset just benefit the largest noncommunity banks and that large size, data that deviated significantly from those of other survey community banks may have had an advantage over their respondents (the reported value was less than or greater than the reported values for 99 percent of responders) were modified to equal smaller peers when deciding to adopt technology. the value reported by a community bank at the 1st or 99th percentile . This was done to limit the effect of outlying data without removing it 8 Jacewitz, Kravitz, and Shoukry (2020) . completely .

6-6 FDIC Community Banking Study ■ December 2020 Box 6.2 Ways That Banks Obtain Technology

Banks obtain new technology in a number of ways. They build it in-house, buy it through merger and acquisition or direct investment, “rent” it by contracting with outside providers including core service providers, or share in it by partnering with other financial and nonfinancial institutions. These pathways are not new, yet much is unknown about the extent to which community banks use each approach.

Data from the CSBS survey indicate that community banks seldom build or buy technology for use in-house. Over three-quarters (78 percent) of community banks participating in the survey responded that they “rarely” or “never” relied on in-house technology for non-lending digital banking products and services (Chart 6.2.1). Of the 218 community banks that offered at least one lending-related technology, almost three-quarters (73 percent) responded that they “rarely” or “never” relied on in-house technology for online loan products. Responses did not vary significantly by adoption category (or, in the case of lending-related technologies, there were too few low-adopting banks for any distinctions to be drawn).

Chart 6.2.1

Percentage of Community Banks Relying on In-House Technology, 2019

Non-Lending Digital Banking Products Online Loan Products

Always Always Usually 4% 6% 10% Usually 14% About Half the Time 8% About Half the Time 7% Never Never 41% 46% Rarely Rarely 32% 32%

Source: Conference of State Bank Supervisors.

In contrast, 94 percent of community banks in the CSBS study had relationships with outside providers of digital banking products and services. Among respondents with at least one such relationship, 41 percent of high-adopting community banks sought to expand those relationships, compared with 39 percent of medium-adopting banks and 24 percent of low-adopting banks.

The frequent use of outside technology service providers suggests that further research into these relationships could deepen the understanding of how community banks obtain technology and may reveal additional factors that influence technology adoption. Future work should include the role of core service providers and should attempt to discover whether the challenges expressed by community banks and referenced briefly in this chapter are exceptions, or may be associated with broader differences in technology adoption. As stated by one community-bank executive, “We are currently captive to our core provider and can only move as fast as they are willing to go. We have many initiatives (e.g., debit card tokenization) that are effectively stalled while we wait for [core service provider].”

Future work could also consider whether and how assistance from external sources—for example, shared innovation labs and accelerators, such as the Alloy Labs Alliance and the ICBA ThinkTECH Accelerator—has facilitated community banks’ adoption of technology.

FDIC Community Banking Study ■ December 2020 6-7 Chart 6.3

Percentage of Community Banks in Each Technology-Adoption Category by Asset Size, First Quarter 2019

Low Medium High

Less Than $100 Million

$100 Million to $200 Million

$200 Million to $500 Million

$500 Million to $1 Billion

More Than $1 Billion

0% 25% 50%75% 100%

Sources: FDIC and Conference of State Bank Supervisors.

Table 6.2 Adoption Rates for the Largest and Smallest Revenue is one indicator of the ongoing resources that a Community Banks community bank may have available if it is to invest in Less Than More new technology. While highly correlated with asset size, $100 Than $1 Million All Billion revenue may be used as a separate measure to account Online Loan Applications 15 .8 37 2. 60 .6 for banks that earned higher yields on their assets or substantial fee income, which banks would be able to Online Loan Closing 3 .2 4 .8 6 .1 direct toward technology. When taken as a share of assets, Mobile Banking 62 .1 90 9. 100 .0 total revenue was, on average, 0.3 percentage points Electronic Bill Payment 65 .3 83 2. 89 .4 greater for high-adopting banks than for low-adopting Automated Loan Underwriting 4 .2 11 6. 33 .3 banks (for a discussion of net income, see section below Interactive Teller Machines 2 .1 10 8. 21 .2 on performance). When high-adopting banks with Remote Deposit Capture 45 .3 78 8. 98 .5 between $100 million and $200 million in total assets are Sources: FDIC and Conference of State Bank Supervisors . compared with low-adopting banks of the same size, the high-adopting banks earned 16 percent more revenue (in of community banks with assets of more than $1 billion dollars) than low-adopting banks.10 (Chart 6.3). Similarly, the adoption rate for each of the seven technology-enabled products and services among the smallest community banks was below the comparable “Our budget will never compete with larger banks’ rate for all community banks in the survey. The opposite budgets.” was true for banks with assets of more than $1 billion —(Medium-adopting) community-bank executive (Table 6.2). “Small banks do not have the resources to implement and manage new and upcoming technologies. We Community Banks With Higher Revenue Were must wait until the products have been implemented Also Greater Technology Adopters by others and proven to be acceptable from a cost and risk standpoint.” To adopt new technology, banks of all sizes require —(Low-adopting) community-bank president resources, including staff, knowledge, time, and funding. To the extent that the costs of these resources take up a greater portion of available budgets, community banks 10 There were roughly equal numbers of low-adopting and high- may be less willing or less able to adopt technology adopting community banks with between $100 million and compared with banks with fewer resource constraints. $200 million in total assets .

6-8 FDIC Community Banking Study ■ December 2020 Chart 6.4

Total Assets and Revenues in 2015 by Change in Adoption Status Between 2015 and 2019 Total Assets Total Revenue $ Millions $ Millions 400 20

300 15

200 10

100 5

0 0 Mobile Online Remote Electronic Mobile Online Remote Electronic Banking Loan Deposit Bill Banking Loan Deposit Bill Applications Capture Payment Applications Capture Payment

O ered Technology by 2019 Did Not O er Technology by 2019

Sources: FDIC and Conference of State Bank Supervisors.

Bank Size and Resources May Have Influenced while the second group (gold bars) reported no change Technology Adoption, or Been Influenced by It, (i.e., the bank did not offer the technology.) 11 For the four or Both technologies included in the survey every year, banks that changed their adoption status and began to offer Of the factors examined in this chapter, size and the technology had, on average, higher assets and higher resources—as measured by assets and revenues—had revenue in 2015 (before adoption). the greatest ties to technology adoption. This naturally raises the question of whether and how much these factors As discussed in the next section, compared with other predated banks’ adoption of technology or whether they community banks in the survey, the 2019 cohort of arose afterward. For example, a larger bank may have low-adopting banks has also experienced slower asset been more likely to adopt technology because of the lower growth in each year from 2015 to 2018. However, without marginal costs associated with economies of scale. It is additional data, it is unclear whether these differences also possible that the bank used technology to expand its existed before technology adoption, or whether the offerings and enter into new markets, leading to increased adoption of technology increased asset growth, or both. size and revenues through growth.

This question is hard to answer with the data available, yet Other Bank Characteristics Were Also there is at least some evidence that differences in asset size Associated With Technology Adoption and revenues predated, and thus potentially influenced, While community banks that adopted technology were community banks’ technology adoption decisions. As most distinguishable by their larger size and higher mentioned above, CSBS has conducted a survey in each year revenues, other characteristics identified in the research since 2015. Although the same banks did not participate and survey responses were also associated with technology in each survey, some overlap existed between years. adoption. Chart 6.4 compares two groups of community banks that participated in either the 2015 or 2016 survey and reported 11 that their bank did not offer a particular technology Of community banks that reported their adoption status in 2015 or 2016 and again in 2018 or 2019, 78 banks did not offer electronic bill product or service at that time. When the same banks were payment in the earlier period, 96 did not offer mobile banking, 114 did surveyed again in 2018 or 2019, the first group (indicated not offer remote deposit capture, and 235 did not offer online loan by the blue bars in Chart 6.4) reported a change in their applications . By 2019, 60 had adopted electronic bill payment (18 had not), 70 had adopted mobile banking (26 had not), 64 had adopted adoption status (i.e., the bank offered the technology), remote deposit capture (50 had not), and 65 had adopted online loan applications (170 had not) .

FDIC Community Banking Study ■ December 2020 6-9 community banks to report higher loans to assets than “[Most promising opportunity is to] expand low-adopting banks. commercial deposit and commercial loan growth.” - (Low-adopting) community-bank president As expected, among community banks in the CSBS survey, technology adoption was associated with “Bank is at historically high loan volumes and historically high loan commitments. New technology higher shares of loans to assets. Chart 6.5 shows that can help overhead from not increasing too much.” high‑adopting banks held, on average, 10 percent more - (Medium-adopting) community-bank executive loans as a share of assets than did low-adopting banks. A higher proportion of loans to assets was not associated with any single technology. Comparing the individual Total Loans Mattered More Than Loan Type offerings among all community banks in the survey, one sees that for each of the lending-related technologies— Loans constitute about two-thirds of a typical community online loan applications, online loan closure, and bank’s assets. Technology offers an opportunity to build on automated loan underwriting—banks that offered a and improve this critical function by increasing the speed product or service (indicated by the light blue bars in and convenience of the application process and producing Chart 6.5) had a higher share of loans to assets than those faster underwriting decisions. Community banks with that did not (gold bars). larger loan books may find these benefits more attractive, compared with their costs, than banks with fewer loans. As also shown in Chart 6.5, high-adopting banks held Technology may also allow banks to increase their lending a greater percentage of their assets in residential loans through new and expanded products and entry into new and C&I loans, and a lesser percentage of their assets in markets. In both cases, we would expect high-adopting consumer loans, than did low- and medium-adopting

Chart 6.5

Total Loans as Share of Assets by Technology-Adoption Category and Product or Service Oering, First Quarter 2019 Total Loans Percent of Assets 80.4 O ered Product or Service Did Not O er Product or Service 73.5 Group Mean 66.7

59.8

53.0 Low Medium High Online Online Loan Automated Loan Closure Loan Applications Underwriting

Residential Loans C&I Loans Consumer Loans Percent of Assets Percent of Assets Percent of Assets 36.8 14.3 4.9

29.3 11.4 3.7 Group Group Group Mean Mean Mean 21.8 8.6 2.5

14.3 5.8 1.2

6.8 3.0 0.0 Low Medium High Low Medium High Low Medium High

Sources: FDIC and Conference of State Bank Supervisors. Note: To indicate the significance of the di erences between the technology adoption categories, the y-axis of each chart was scaled to be roughly one standard deviation below and above the mean for all community banks in the survey.

6-10 FDIC Community Banking Study ■ December 2020 Chart 6.6

Technology Adoption by Lending Specialization Technology-Adoption Category

All Low Medium High

Agricultural

Mortgage

CRE

C&I

Does bank o er online loan applications? Yes No

All

Agricultural

Mortgage

CRE

C&I

Does bank use automated loan underwriting? Yes No

All

Agricultural

Mortgage

CRE

C&I 0% 25% 50% 75%100%

Sources: FDIC and Conference of State Bank Supervisors. Note: Appendix A outlines the criteria for each lending specialty.

banks.12 The differences between low- and high-adopting Another way to examine whether technology adoption banks—0.7 percent of assets for consumer loans, varied by lending type is to compare community banks 2.2 percent for C&I loans, and 4.9 percent for residential that specialized in certain types of lending. High-adopting loans—were not as large as the difference mentioned banks made up the greatest percentage of C&I specialists, above for total loans to assets. Nonetheless, these findings relative to the other lending specializations (Chart 6.6). suggest possible dissimilarities in the benefits, costs, or These banks were also more likely to have adopted online availability of technology between the three loan types. loan applications and automated loan underwriting, compared with all community banks. If one assumes that community-bank business lending typically involves a more hands-on process, as suggested in Chapter 4, these

12 Residential mortgage lending consists of loans secured by 1–4 findings may be unexpected. However, these results family or multifamily (5 or more) residential properties . Consumer may reflect the use of technology in parts of the lending loans consist of loans to individuals for household, family, and other process (since the portion of the application process that is personal expenditures—for example, credit card loans, student loans, and automobile loans . online or the degree to which underwriting is automated

FDIC Community Banking Study ■ December 2020 6-11 Chart 6.7

Total and Core Deposits as Share of Assets by Technology-Adoption Category

Total Deposits Core Deposits

Percent of Assets Percent of Assets 89.3 85.3

86.6 81.1

Group Group Mean Mean 84.0 77.0

81.3 72.9

78.7 68.8 Low Medium High Low Medium High

Sources: FDIC and Conference of State Bank Supervisors. Note: To indicate the significance of the dierences between the technology adoption categories, the y-axis of each chart was scaled to be roughly one standard deviation below and above the mean for all community banks in the survey.

was not specified by survey respondents). Or the results as a share of assets did not vary widely by technology- may also reflect increased competition from nonbanks, adoption category (Chart 6.7). For low-adopting banks, as indicated by a 2020 study that found small businesses deposits as a share of assets was less than a percentage were 12 percentage points more likely to receive financing point higher relative to medium- and high-adopting through a fintech or online lender in 2018 than in 2016, banks. Core deposits, which make up the bulk of with a nearly equal decline in the likelihood of borrowing community-bank deposits, were slightly favored by from a bank lender.13 low-adopting and medium-adopting banks relative to high-adopting banks; when measured as a share of In contrast, high-adopting banks were least represented deposits, however, core deposits varied by less than among agricultural specialists. Such a result is not one-half of 1 percentage point between the technology surprising, given that agricultural specialists tend to be adoption categories.14 Even for the individual product smaller and therefore (as previously indicated) less likely and service offerings, results were mixed. Shares of total to adopt technology. Agricultural lending may also be more deposits and core deposits were higher for community specialized, making automation and online processes less banks that adopted mobile banking and electronic bill effective or harder to implement. payment but were lower for banks that adopted remote deposit capture and ITMs. Technology Was Not Associated With Deposits Low-Adopting Banks Generally Had Slower Growth Community banks fund most of their assets with in Assets and Deposits deposits, and banks in the CSBS survey were no exception: in first quarter 2019, on average, 84 percent As mentioned above, community banks frequently of their assets were funded with deposits. Given the cited customers and customer growth as promising important role of deposit funding, we might expect opportunities that could follow from the adoption of technology, particularly technology that enhances the technology. Therefore, we might expect assets and deposits functionality of deposit accounts, to be more prevalent in to grow faster for banks that adopted technology. institutions with larger ratios of deposits to assets. For community banks in the CSBS survey, however, deposits 14 Core deposits were calculated according to the definition in the Uniform Bank Performance Report—i e. ,. as the sum of all transaction 13 The study uses the terms “fintech lender” and “online lender” accounts, nontransaction money-market deposit accounts (MMDAs), interchangeably to refer to any nonbank online lender, as reported nontransaction other savings deposits (excluding MMDAs), and in the Federal Reserve’s Small Business Credit Survey . Barkley and nontransaction time deposits of $250,000 and less, minus fully Schweitzer (2020) . insured brokered deposits of $250,000 and less .

6-12 FDIC Community Banking Study ■ December 2020 “We have a lot of room for growth and improvement with new technology.” —(High-adopting) community-bank executive

“[Most promising opportunity regarding new technology is] [m]arket opportunity to increase market share by expanding banking services [and] by utilizing ITMs to control cost of doing so.” —(High-adopting) community-bank president

Chart 6.8 low-adopting banks. Over the same four quarter period, Distribution of Year-Over-Year Growth in Assets and deposits in high-adopting banks grew by an average of Deposits by Technology-Adoption Category, 6.1 percent, slightly more than the average for medium- Interdecile and Interquartile Ranges adopting banks (5.9 percent) and significantly more than Low the 3.4 percent growth experienced by low-adopting banks. Assets Medium High The difference between low adopters and high adopters was most pronounced at the lower ends of the growth distribution (Chart 6.8). With respect to assets, high- Deposits adopting banks had significantly higher growth at the 10th and 25th percentiles, growing 0.2 percent and 1.9 percent, -10 010 20 Percent respectively, compared with -4.7 percent and -0.5 percent Sources: FDIC and Conference of State Bank Supervisors. for low-adopting banks. Similarly, for deposits, high- Note: Lighter shades indicate the interdecile range (10th to 90th percentile); adopting banks at the 10th and 25th percentiles grew by darker shades indicate the interquartile range (25th to 75th percentile). -0.9 percent and 0.8 percent, respectively, which was much higher than the -5.7 percent and -1.8 percent growth Among community banks participating in the CSBS study, experienced by low-adopting banks. high-adopting banks experienced higher average growth in both assets and deposits relative to medium- and As Chart 6.9 shows, the difference in growth between low- low-adopting banks. For high-adopting banks, asset and high-adopting banks did not appear transitory. From growth from the first quarter of 2018 to the first quarter 2015 to 2018 low-adopting banks, as defined in 2019, grew of 2019 was 6.3 percent, on average, compared with their assets between 1.7 and 3.9 percentage points slower 6.1 percent for medium-adopting banks and 4.4 percent for

Chart 6.9

Annual Growth in Assets and Deposits by 2019 Technology-Adoption Category, 2015–2018

Assets Deposits Percent Percent 8.0 8.0

6.0 6.0

4.0 4.0

2.0 2.0

0.0 0.0 2015 2016 2017 2018 2015 2016 2017 2018 Low Medium High

Sources: FDIC and Conference of State Bank Supervisors.

FDIC Community Banking Study ■ December 2020 6-13 than high-adopting banks. For deposits, the difference in year-over-year growth between the two groups ranged “Utilizing new technologies also helps to improve between 1.9 percent and 4.3 percent over the same period. productivity and efficiencies, which are necessary in There was no consistent pattern in the difference in asset order to remain profitable and competitive.” and deposit growth between medium- and high-adopting —(Low-adopting) community-bank president banks from 2015 to 2018. Although growth during 2015 “We are excited to look into the AI platforms and see and 2016 favored high-adopting banks, medium-adopting how this can help our bank’s profits and reduce our banks outpaced high-adopting banks in 2017 for both salary expenses.” assets and deposits and in 2018 for assets. —(Low-adopting) community-bank president High-Adopting Community Banks Outperformed Other Banks in the Survey, but the Reasons income, compared with 95.4 percent for low adopters. Were Unclear High adopters reported annual growth in net income that Performance may be associated with technology was 7.7 percentage points higher than the comparable adoption to the extent that it indicates a greater or lesser reported growth of low adopters, and nearly 8 percent capacity for the bank to invest in technology or if banks more high‑adopting banks increased their earnings from that adopt technology become more efficient or more the previous year. Differences between medium- and adept at marketing or pricing products and services. high‑adopting banks followed a similar pattern but were Table 6.3 shows that high-adopting community banks smaller in magnitude. Compared with high adopters, in the CSBS survey were more likely to be profitable and 0.5 percent fewer medium-adopting banks were profitable experience earnings gains in 2018, relative to low- and and 1.3 percent fewer experienced earnings gains in 2018. medium‑adopting banks. High adopters earned a pre-tax return on average assets that was 21 basis points greater Comparing the components of return on assets, it appeared than the return of low-adopting banks, on average, with that noninterest income was mainly responsible for the 99.5 percent of high adopters generating positive net higher returns experienced by high-adopting banks. In

Table 6.3 Average Performance Measures by Technology-Adoption Category, 2018

All Low Medium High Net Income (Pretax), 2018: Percent With Positive Net Income (Profitable) 98 .3 95 .4 99 .0 99 .5 Year-Over-Year Growth, Percent 22 .1 19 .3 19 .0 27 .0 Percent of Institutions With Earnings Gains 74 .9 69 .5 76 .0 77 .3 Percent of Average Assets 1 .23 1 .09 1 .27 1 .30 Components of Return on Assets (Percent of Average Assets) Interest Income 4 .13 4 .07 4 .20 4 .09 Interest Expense 0 .57 0 .56 0 .58 0 .56 Noninterest Income 0 .68 0 .54 0 .62 0 .83 Service Charges on Deposit Accounts 0 .18 0 .18 0 .20 0 .16 Noninterest Expense 2 .93 2 .91 2 .86 3 .02 Expenses for Salaries and Benefits 1 .69 1 .62 1 .66 1 .76 Cost of Earning Assets (bp) 62 61 63 61 Net Interest Margin (bp) 389 382 396 388 Average Cost of Interest-Bearing Deposits (bp) 74 75 76 72 Efficiency Ratio 68 .3 69 .8 67 .4 68 .1 Sources: FDIC and Conference of State Bank Supervisors . Notes: Basis point (bp) = 1/100th of 1 percent; efficiency ratio is equal to noninterest expense as a share of operating income .

6-14 FDIC Community Banking Study ■ December 2020 2018, noninterest income as a percentage of average assets was 29 basis points higher for high-adopting banks than “Being in a more rural area, customers don’t require for low-adopting banks and 21 basis points higher than the newest technology as soon as other areas and for medium-adopting banks. This difference, however, there is less local competition.” was not associated with higher service charges on deposit —(Low-adopting) community-bank president accounts, as earlier research on transactional websites had suggested, and instead was attributable mainly to “other defined by a bank’s physical location may no longer apply noninterest income.” in the same manner as it has in the past.15

For high-adopting banks compared with other banks in Differences between urban and rural consumers in their the survey, the higher return associated with noninterest demands and capabilities may affect a community bank’s income was partially offset by a higher ratio of noninterest decision to adopt or not adopt technology. For example, expense to average assets. The difference in noninterest a “digital divide” between rural and urban Americans expense largely arose because of a 10 to 14 basis point has been documented for many years, with 2019 data differential in expenses for salaries and employee benefits. from the National Telecommunications and Information Higher staff costs for high-adopting banks contradicts the Administration showing a 6 percentage point differential argument that technology—specifically, automation— between urban and rural areas in the share of people reduces staff time devoted to manual processes but using the internet at home. This difference increased to coincides with the theory that banks use technology as a 8 percentage points for smartphone use.16 Survey data complement to, rather than a substitute for, human capital. collected in 2017 by the Pew Research Center found that It is also possible that more specialized and potentially rural adults were less likely to have multiple devices with more expensive expertise was needed to adopt technology, internet access, less likely to use the internet on a daily resulting in higher costs for salaries and benefits for basis, and more likely to never go online, compared with high‑adopting banks relative to low- and medium- suburban and urban counterparts.17 adopting banks.

Table 6.4 shows that among community banks in the CSBS There were minimal differences in interest income survey, the probability of being a low-technology adopter and interest expense between the adoption categories. increased from 28 percent to 39 percent if the bank was Similarly, technology adoption did not appear to bear located in a rural area (defined in the data as “other area”). any relationship to cost of earning assets, net interest Conversely, the probability of being a low-technology margin, average cost of interest-bearing deposits, or adopter decreased from 51 percent to 43 percent if the efficiency ratio. This may be because the technologies bank was located in an urban area (defined in the data as included in the survey did not translate to differences in “metropolitan area”). The higher share of low adopters these measures, or it may be because any differences have among rural community banks persisted even after not yet materialized. As one community-bank president differences in asset size were accounted for. The opposite said, “In the short term, it [technology] does not improve pattern was true for the likelihood that a community bank the efficiency ratio, but in the long term the bank may be was a high adopter, although in this case, for banks of rewarded by the retention of younger customers and the similar asset size, location in a rural or urban area had less future business opportunities they may provide.” of an effect.

Environmental Factors Were Linked to Community banks in areas with low population or Technology Adoption economic growth may be less likely to invest in technology if those banks are concerned that slow growth will The environment a community bank operates in can limit their future revenue or customer base. Similarly, affect customer demand, the ability to hire employees, and current and future resources, all of which may play a 15 In this chapter, unless otherwise specified, environmental factors role in a bank’s decision of whether to adopt technology. were measured on the basis of the location of a community bank’s On the other hand, with the power to connect banks and main office . 16 National Telecommunications and Information Administration customers virtually, the concept of “environment” as (2020) . 17 Perrin (2019) .

FDIC Community Banking Study ■ December 2020 6-15 Table 6.4 Characteristics of Bank Environment by Technology-Adoption Category

Low- Medium- High- All Banks in Adopting Adopting Adopting Survey Banks Banks Banks Main Office Location (Percent in Each Category): Metropolitan Area (Urban Area) 51 .1 42 .7 48 .7 59 .0 Micropolitan Area 21 .0 18 .3 20 .2 23 .6 Other Area (Rural Area) 27 .9 38 .9 31 .1 17 .4 Population Growth: Cumulative Annual Growth From 2010 to 2018, Percent 0 .22 0 .13 0 .19 0 .31 Located in a Depopulating County (2010 to 2018) 42 .2 51 .1 40 .4 37 .9 Median Age of Local Population (2018) In Years 39 .9 40 .8 39 .3 39 .9 Located in County in the Highest (Oldest) Quartile 25 .4 36 .6 18 .1 25 .1 Located in County in the Lowest (Youngest) Quartile 22 .7 16 .8 26 .4 23 .1 Cumulative Annual GDP Growth From 2010 to 2018 (Percent) 3 .09 2 .85 3 .04 3 .31 “Greatest Single Challenge” Facing Bank is Business Conditions 7 .1 9 .7 5 .8 6 .7 Average Competitors Within 10 Miles (Percent in Each Category): Less Than 2 8 .1 16 .8 7 .8 2 .6 2 to 5 24 .5 32 .1 26 .4 17 .5 5 to 10 28 .6 29 .8 25 .4 30 .9 10 to 25 29 .9 16 .0 32 .1 37 .1 More Than 25 8 .9 5 .3 8 .3 11 .9 Share of Deposits Within 10 Miles (Percent in Each Category): Less Than 10 Percent 30 .2 24 .4 32 .8 31 .4 10 Percent to 33 Percent 44 .7 37 .4 43 .8 50 .5 More Than 33 Percent 25 .1 38 .2 23 .4 18 .0 “Greatest Single Challenge” Facing Bank is Competition 14 .9 12 .4 12 .7 19 .0 Sources: FDIC, Conference of State Bank Supervisors, and Bureau of Economic Analysis . Note: Counties in the youngest 25 percent are those where the median age is 36 .6 years or below; counties in the oldest 25 percent are where the median age is 42 .5 years or above (see Chapter 3 in this study) . banks may be less likely to prioritize technology if they 2018 the average county‑level CAGR for high-adopting are located near fewer customers who demand or use banks was 3.3 percent, compared with 3.0 percent for it—for example, areas with a higher median age.18 On the medium-adopting banks and 2.9 percent for low-adopting other hand, such banks may also be motivated to adopt banks (Table 6.4). When banks of similar asset size technology to expand into growing markets or to attract were compared, the difference between high- and and retain younger customers, as indicated by multiple low-adopting banks narrowed slightly but did not community banks in the CSBS survey.

“Mobile deposit has helped retain some of our younger On average, high-adopting community banks in the customers as they go off to the big cities to college.” CSBS survey were located in counties with higher —(Low-adopting) community-bank president economic growth, as measured by the cumulative annual growth rate (CAGR) for GDP. Between 2010 and “Another challenge is to persuade the senior generations (baby boomers my age and older) to 18 Vogels (2019) . Another online survey conducted in 2019 found accept and utilize the new technology.” that 62 percent of those ages 18 to 29 banked using a mobile app compared with 22 percent of those ages 55 to 64 and 7 percent of —(Low-adopting) community-bank president those 65 and older . American Bankers Association (2019) .

6-16 FDIC Community Banking Study ■ December 2020 Chart 6.10

Number of Local Competitors and Share of Deposits by Technology-Adoption Category Number of Competitors Share of Deposits Within 10 Miles

Low 7.7 30.3% 10 Miles

Medium 11.1 23.3% 10 Miles

High 13.5 19.9% 10 Miles

Within 5 Miles Within 10 Miles Sources: FDIC and Conference of State Bank Supervisors. Note: “Competitor” is defined as any bank (including noncommunity banks and community banks not included in the CSBS survey) that operated a branch within 5 or 10 miles. Because community banks may operate branches in multiple areas, number of competitors and share of deposits were calculated as the simple average of all full-service branches operated by the bank as of June 30, 2019.

disappear. Low-adopting banks were also more likely to cite “business conditions” as the greatest single challenge “We invest in and use technology because the market facing their bank. The pattern did not hold for all adoption place requires us to do so.” categories, however, since high-adopting banks were more —(High-adopting) community-bank executive likely to cite this challenge than medium-adopting banks. “New technologies of every kind offer our bank a better opportunity to stay competitive with the large Other local factors, such as median age and population regional banks and the money-center banks.” growth, did not have a strong tie to technology adoption. —(Medium-adopting) community-bank executive From 2010 to 2018, while low-adopting banks were more often located in counties with negative population growth market—as measured by the number of banks (including (51 percent, as opposed to 40 percent for medium-adopting noncommunity banks) operating within a certain distance banks and 38 percent for high-adopting banks) and with and by the share of local deposits held by the bank—to a slower average CAGR, these differences disappeared play a role in technology adoption. Community banks with after asset size was accounted for. Similarly, when banks a larger share of local deposits or that operate in close of similar size were compared, differences in the average proximity to fewer banks would likely feel less pressure median age and the share of banks located in the youngest to adopt new technology, compared with banks that have and oldest counties by quartile declined in magnitude. a smaller share of deposits and a greater number of local competitors. Responses to the CSBS survey indicate that competition was a consideration for community banks, with most Using data on deposits and location by branch from respondents viewing banks located within their market the FDIC’s Summary of Deposits survey, low-adopting as their greatest source of competition.19 Therefore, we banks tended to operate in markets with fewer average might expect the level of competition within a bank’s competitors per bank branch. (For this chapter, a community bank’s market was the five- and ten-mile 19 Over 15 percent of CSBS survey respondents (including those not radius surrounding each of the bank’s branches.) Using examined in this chapter) selected competition as the “single greatest this measure, low-adopting banks faced an average challenge” facing their bank; only core deposit growth (22 percent) and regulation (16 percent) registered more responses . For all but of 7.7 competitors, compared with 11.1 competitors two products and services (wealth management/retirement services for medium-adopting banks and 13.5 competitors for and payment services), over 75 percent of respondents indicated that their greatest source of competition came from institutions with a high‑adopting banks (Chart 6.10). Low-adopting banks headquarters, a branch, or a satellite office in their market . also operated in markets where they held 10.4 percent

FDIC Community Banking Study ■ December 2020 6-17 more deposits as a share of all the deposits held by banks within ten miles, compared with high-adopting banks. “Community banks have the opportunity to show This difference decreased only marginally after bank size and prove to customers that their technology can was accounted for. rival that of much larger banks.” —(Medium-adopting) community-bank executive

Attitudes Toward Technology and “Community banks survive on quality, personal Expectations About Profitability and customer service, not technology advancements.” Expansion Played a Role in Adoption Decisions —(Low-adopting) community-bank president Not surprisingly, technology adoption differed depending on the importance a bank attributed to technology. For and its environment to include, in addition, the bank’s example, 81 percent of high-adopting banks responded attitudes toward technology. that technology adoption was either “important” or “very important,” compared with 71 percent of medium- Further, technology adoption varied by attitudes and adopting banks and 56 percent of low-adopting banks. expectations not directly related to technology. Chart 6.11 With respect to technology leadership, 32 percent of displays responses to four questions about a bank’s high-adopting banks felt that it was “important” or “very expectations for business conditions, profitability, and important” to be a leader in new or emerging technologies, regulatory burden over the next 12 months, as well as compared with 28 percent for medium-adopting banks the bank’s overall outlook for expansion. In each case, and 14 percent for low-adopting banks. The fact that most low-adopting banks tended to have more-pessimistic banks, including high adopters, stated that technology views than did medium-and high-adopting banks. The adoption—but not technology leadership—was important largest differences appeared in expectations for future aligns with the analysis above indicating that community profitability and outlook for expansion. The percentage of banks were generally focused on “keeping up” rather low adopters that believed profitability would be higher than leading in technology. The findings also suggest that over the next 12 months trailed the percentage of high- technology adoption goes beyond a bank’s characteristics

Chart 6.11

Expectations of Bank Leadership by Technology-Adoption Category Better Same Worse How do you expect High business conditions in Medium your market to change over the next 12 months? Low

Higher Same Lower How do you expect your High profitability to change Medium over the next 12 months? Low

Yes No Overall, is the current High period a good time to Medium expand your bank’s operations? Low

Lighter Same Heavier How do you expect the High regulatory burden on Medium your bank to change over the next 12 months? Low 0255075100 Percent of Responders Sources: FDIC and Conference of State Bank Supervisors.

6-18 FDIC Community Banking Study ■ December 2020 adopting banks by nearly 19 percentage points and trailed help us not only identify changes in adoption patterns medium-adopters by 27 percentage points. Just over half over time but also incorporate new technologies as they of low-adopting banks believed that the current period (at become available. The former is particularly relevant given that time, spring 2019) was a good time to expand bank the short- and long-term changes in technology use and operations, compared with 81 percent of medium-adopting adoption that may arise from the COVID-19 pandemic banks and 87 percent of high-adopting banks. (see Box 6.3).

Future Research Will Yield Greater Insights Third, further research may explore whether the factors Into Technology Adoption explored in this chapter, as well as others, may affect the decisions of different subsets of community banks To explore how banks that adopted technology differed to adopt or not adopt technology. Such work may also from those that did not, this chapter has examined several help inform policy discussions on other topics—for characteristics of community banks, their environment, example, deposit flight from depopulating rural areas. and the attitudes and expectations of their leadership. A comparison of the technology profiles of community For respondents to the CSBS survey, size and revenue banks located in rural areas with a declining population were the main factors differentiating low adopters, could help determine whether certain technologies helped medium adopters, and high adopters among community some banks in those areas retain customers or attract banks. Other factors, including a bank’s expectations and out-of-market deposits. attitudes toward technology, its ratio of loans to assets, and its competitive environment, were also relevant, but Finally, future research should incorporate data from not as much as size and revenue. all community banks to the greatest extent possible. While community banks participating in the CSBS survey In the future, the FDIC plans to undertake additional generally reflected the wider population, any differences research to overcome some of the limitations of this could prevent the broader application of findings reported chapter. First, the measure used to differentiate “low,” in this chapter. For example, the CSBS survey did not “medium,” and “high” adopting banks cannot account for include responses from community banks with national the length of time that a community bank offered or used charters or from banks in every state. Such banks may a particular technology or for the quality and functionality approach their technology adoption decisions differently; of the technology. In the future, such information could therefore, it would benefit future researchers if these be collected and analyzed to determine whether specific differences were eliminated as much as possible. components or uses of technology were associated with the factors studied here and whether these associations varied As the primary federal regulator for most community by early and late adopters. banks, the FDIC encourages further research into factors that may have influenced, or may have been influenced Second, while these findings include some evidence of by, a community bank’s technology adoption decisions. directional effects, data collected over a longer period The FDIC also encourages further research in the use may help us distinguish between two types of effect: the of technology in community banks in general. Ongoing effects of different factors on a bank’s decision to adopt research and data collection is needed to keep pace with technology, and the effects of adopting technology on rapidly evolving technology and to better understand the those factors. In addition, ongoing data collection will benefits and risks of community banking in a digital age.

FDIC Community Banking Study ■ December 2020 6-19 Box 6.3 Technology Adoption and the COVID-19 Pandemic

The COVID-19 pandemic has given rise to a defining change for community banks: a broader use of technology, both at present and for the future. The pandemic has resulted in branch closures, stay-at-home orders, and a general desire to limit direct contact, all of which has increased the use of computers, mobile phones, and other smart devices to complete financial services transactions. To meet growing demand, community banks have used both direct investment and contracts with technology service providers and fintechs to accelerate their adoption of technologies that enable such services as remote deposit, online applications, peer-to-peer payments, and electronic signatures.

Some community banks, for example, used technology to help manage the unprecedented volume of loan applications received in response to the Small Business Administration’s (SBA) Paycheck Protection Program (PPP).a Over the span of a few months, community banks provided billions of dollars of needed credit to small and medium-sized enterprises through the program, with 3,843 community banks holding over $148 billion in PPP loans as of June 30, 2020. Arguably, technology facilitated this lending by allowing some community banks to accept applications and supporting documentation online, process applications faster, and submit files for SBA approval.b As the PPP moves into its next phase, community banks are also seeking the aid of technology to automate loan forgiveness applications.c

Not all accounts from community bankers and borrowers about using technology to assist with PPP lending were positive, however, nor is it clear that technology increased the use or efficiency of the program. Reports of difficulties connecting with SBA’s systems (E-Tran) and last-minute changes to the program, including a ban on robotic data entry systems, suggested a limit to the effectiveness of technology.d Nonetheless, at least among community banks in the 2019 CSBS survey, those identified in this chapter as high-technology adopters showed greater participation in the program, with PPP loans totaling 6.5 percent of assets, compared with 5.7 percent of assets for medium-adopting banks and 5.0 percent of assets for low-adopting banks. Future research may better identify the extent to which technology facilitated PPP lending as well as other credit during the pandemic.

The degree to which banks continue after the pandemic to rely on technology investments and partnerships made during the pandemic remains unknown; however, it seems unlikely that customers’ use of technology will return to pre-pandemic levels even after branches and the economy resume normal operations. In a PriceWaterhouseCoopers (PwC) survey of 6,000 U.S. bank customers conducted in May and June 2020, 24 percent stated they were less likely to use their bank’s branch offices. In addition, following months of remote work, banks (like many other businesses) may consider permanent changes to workspaces, which could have long-term effects on branch structure and operating expenses.

It is also possible that because of the pandemic, technology adoption by community banks will decrease. Banks that experience financial hardship may have reduced ability and desire to invest in new technology, a development suggested by the findings of this chapter associating revenues and local economic growth with technology adoption. And post-pandemic, some community banks may experience less of a decline to branch traffic, a development suggested by the number of respondents to the PwC survey who indicated they were likely to continue using branch offices, including for services that can be done remotely.

a As discussed in previous chapters, the PPP provided a federal guarantee for low-interest forgivable loans made to eligible businesses by bank and nonbank lenders . b Groenfeldt (2020) . c Cross (2020) . d Price (2020) .

6-20 FDIC Community Banking Study ■ December 2020 Bibliography

Executive Summary

Federal Reserve Banks. “Small Business Credit Survey: 2020 Report on Employer Firms.” April 2020. https://www.fedsmallbusiness.org/medialibrary/FedSmallBusiness/files/2020/2020-sbcs-employer-firms-report.pdf.

Chapter 2: Structural Change Among Community and Noncommunity Banks

Adams, Robert, and Jacob Gramlich. “Where Are All the New Banks? The Role of Regulatory Burden in New Bank Formation.” Review of Industrial Organization 48 (2016): 181–208.

FDIC. Community Banking Study. December 2012. https://www.fdic.gov/resources/community-banking/report/2012/2012- cbi-study-full.pdf.

Fronk, Jared. “Core Profitability of Community Banks: 1985–2015.” FDIC Quarterly 10, no. 4 (2016): 37. https://www.fdic.gov/ bank/analytical/quarterly/2016-vol10-4/fdic-v10n4-3q16-quarterly.pdf.

National Bureau of Economic Research. Business Cycle Dating. Accessed July 16, 2020. https://www.nber.org/research/ data/us-business-cycle-expansions-and-contractions.

Sullivan, Lauren, and Maria Tor. “Wisconsin Banks Terminate Deal Following 30% Sell-Off.” S&P Global Market Intelligence, May 21, 2020.

U.S. Government Accountability Office. “Community Banks: Effect of Regulations on Small Business Lending and Institutions Appears Modest, but Lending Data Could Be Improved.” GAO–18–312, September 2018. https://www.gao.gov/ reports/GAO-18-312/.

Chapter 3: The Effects of Demographic Changes on Community Banks

Anderlik, John M., and Richard D. Cofer Jr. “Long-Term Trends in Rural Depopulation and Their Implications for Community Banks.” FDIC Quarterly 8, no. 2 (2014). https://www.fdic.gov/bank/analytical/quarterly/2014-vol8-2/article2.pdf.

FDIC. Community Banking Study. December 2012. https://www.fdic.gov/resources/community-banking/report/2012/2012- cbi-study-full.pdf.

FDIC. “Trends in Community Banking.” FYI: An Update on Emerging Issues in Banking, May 18, 2004. https://www.fdic.gov/ bank/analytical/fyi/051804fyi.pdf.

U.S. Census Bureau. Migration data. Accessed October 2020. https://www2.census.gov/programs-surveys/popest/ datasets/2010-2018/counties/totals/co-est2018-alldata.pdf.

—— . American Community Survey. 2010-2018.

Chapter 4: Notable Lending Strengths of Community Banks

American Bankruptcy Institute. September 2020 Bankruptcy Statistics – Commercial Filings. https://www.abi.org/ newsroom/bankruptcy-statistics

Anderlik, John and Jeffrey Walser. “Rural Depopulation: What Does It Mean for the Future Economic Health of Rural Areas and the Community Banks That Support Them?” FDIC Banking Review, 16, no. 3:57-95, 2004.

Anderlik, John and Richard D. Cofer Jr. “Long-Term Trends in Rural Depopulation and Their Implications for Community Banks.” FDIC Quarterly 8, no. 2 (2014). https://www.fdic.gov/bank/analytical/quarterly/2014-vol8-2/article2.pdf.

Board of Governors of the Federal Reserve System. “Financial Accounts of the United States.” Report Z. 1, March 2020.

FDIC Community Banking Study ■ December 2020 i Brennecke, Claire, Stefan Jacewitz, and Jonathan Pogach. “Shared Destinies? Small Banks and Small Business Consolidation.” FDIC Working Paper Series 2020-04, July 2020. https://www.fdic.gov/bank/analytical/cfr/2020/wp2020/ cfr-wp2020-04.pdf.

FDIC. Community Banking Study. December 2012. https://www.fdic.gov/resources/community-banking/report/2012/2012- cbi-study-full.pdf.

Federal Reserve Banks. “Small Business Credit Survey: 2020 Report on Employer Firms.” April 2020. https://www.fedsmallbusiness.org/medialibrary/FedSmallBusiness/files/2020/2020-sbcs-employer-firms-report.pdf.

U.S. Department of Agriculture. Farm Balance Sheet and Financial Ratios, U.S. https://www.ers.usda.gov/data-products/ farm-income-and-wealth-statistics/data-files-us-and-state-level-farm-income-and-wealth-statistics/

U.S. Census Bureau. American Community Survey. 2018.

—— . Current Population Survey/Housing Vacancy Survey. March 10, 2020.

U.S. Small Business Administration. “Gross Approval Amount by Program Report.” July 2020. https://www.sba.gov/ document/report-small-business-administration-loan-program-performance

U.S. Small Business Administration. “Paycheck Protection Program Report Through August 8, 2020.” https://www.sba. gov/document/report-paycheck-protection-program-report-through-august-8-2020

Yelp Economic Average. “Yelp Local Economic Impact Report.” September 2020. https://www.yelpeconomicaverage.com/ business-closures-update-sep-2020

Chapter 5: Regulatory Change and Community Banks

Adams, Robert, and Jacob Gramlich. “Where Are All the New Banks? The Role of Regulatory Burden in New Bank Formation.” Review of Industrial Organization 48 (2016): 181–208.

American Bankers Association. 23rd Annual ABA Residential Real Estate Survey Report. April 2016.

Bhutta, Neil, and Daniel R. Ringo. “Residential Mortgage Lending From 2004 to 2015: Evidence From the Home Mortgage Disclosure Act Data.” Federal Reserve Bulletin 102, no.6 (November 2016).

Brennecke, Claire, Stefan Jacewitz, and Jonathan Pogach. “Shared Destinies? Small Banks and Small Business Consolidation.” FDIC Working Paper Series 2020-04, July 2020. https://www.fdic.gov/bank/analytical/cfr/2020/wp2020/ cfr-wp2020-04.pdf.

Consumer Financial Protection Bureau. 2013 RESPA Servicing Rule Assessment Report. January 2019.

FDIC. Crisis and Response: An FDIC History, 2008–2013. https://www.fdic.gov/bank/historical/crisis/.

Feldman, Ron J., Ken Heinecke, and Jason Schmidt. “Quantifying the Costs of Additional Regulation on Community Banks.” Federal Reserve Bank of Minneapolis Economic Policy Paper, May 30, 2013. https://www.minneapolisfed.org/ article/2013/quantifying-the-costs-of-additional-regulation-on-community-banks.

Fronk, Jared. “Core Profitability of Community Banks: 1985–2015.” FDIC Quarterly 10, no. 4 (2016): 37. https://www.fdic.gov/ bank/analytical/quarterly/2016-vol10-4/fdic-v10n4-3q16-quarterly.pdf.

Jacewitz, Stefan, Troy Kravitz, and George Shoukry. “Economies of Scale in Community Banks.” FDIC Staff Studies, 2020–06, December 2020. https://www.fdic.gov/analysis/cfr/staff-studies/2020-06.pdf.

Shoemaker, Kayla. “Trends in Mortgage Origination and Servicing: Nonbanks in the Post-Crisis Period.” FDIC Quarterly 13 no. 4 (2019). https://www.fdic.gov/bank/analytical/quarterly/2019-vol13-4/fdic-v13n4-3q2019-article3.pdf.

ii FDIC Community Banking Study ■ December 2020 Chapter 6: Technology in Community Banks

American Bankers Association. “Preferred Banking Methods Infographic: How Americans Access Their Bank Accounts.” November 1, 2019. Accessed June 2020. https://www.aba.com/news-research/research-analysis/ preferred-banking-methods.

Barkley, Brett, and Mark E. Schweitzer. “The Rise of Fintech Lending to Small Businesses: Businesses’ Perspectives on Borrowing.” Federal Reserve Bank of Cleveland Working Paper 20-11, 2020.

Conference of State Bank Supervisors. 2019 National Survey of Community Banks. In Community Banking in the 21st Century: Research and Policy Conference. https://www.communitybanking.org/~/media/files/publication/ cb21publication_2019.pdf.

Cross, Miriam. “Banks Seek More Fintech Help for PPP’s Next Phase.” American Banker, July 10, 2020. https://www.americanbanker.com/news/banks-seek-more-fintech-help-for-ppps-next-phase.

Dahl, Drew, Andrew Meyer, and Neil Wiggins. “How Fast Will Banks Adopt New Technology This Time?” The Regional Economist 25 no. 4 (2017).

DeYoung, Robert, William W. Lang, and Daniel L. Nolle. “How the Internet Affects Output and Performance at Community Banks.” Journal of Banking and Finance 31 no. 4 (2007).

Dos Santos, Brian L., and Ken Peffers. “Competitor and Vendor Influence on the Adoption of Innovative Applications in Electronic Commerce.” Information & Management 34 no. 3 (October 1998): 175–184.

Feng, Zifeng, and Zhonghua Wu. “Bank Technology: Productivity and Employment.” 2019.

Groenfeldt, Tom. “Community Banks Met the Challenge of the SBA’s PPP Funding.” Forbes, June 17, 2020. https://www. forbes.com/sites/tomgroenfeldt/2020/06/17/community-banks-met-the-challenge-of-the-sbas-ppp-funding.

Hamilton, Anita. “Banking Goes Mobile.” Time, April 2, 2007. http://content.time.com/time/business/ article/0,8599,1605781,00.html.

Hannan, Timothy H., and John M. McDowell. “The Determinants of Technology Adoption: The Case of the Banking Firm.” RAND Journal of Economics 15 no. 3 (1984).

He, Zhaozhao. “Rivalry, Market Structure and Innovation: The Case of Mobile Banking.” Review of Industrial Organization 47 no. 2 (2015).

Jacewitz, Stefan, Troy Kravitz, and George Shoukry. “Economies of Scale in Community Banks.” FDIC Staff Studies, 2020-06, December 2020. https://www.fdic.gov/analysis/cfr/staff-studies/2020-06.pdf.

Kopchik, Jeffrey. “Remote Deposit Capture: A Primer.” FDIC Supervisory Insights, Summer 2009. https://www.fdic.gov/ regulations/examinations/supervisory/insights/sisum09/sisummer09-article2.pdf.

National Telecommunications and Information Administration. Digital Nation Data Explorer. Accessed June 2020. https://www.ntia.doc.gov/data/digital-nation-data-explorer.

Perrin, Andrew. “Digital Gap Between Rural and Nonrural America Persists.” Pew Research Center, May 31, 2019. https://www.pewresearch.org/fact-tank/2019/05/31/digital-gap-between-rural-and-nonrural-america-persists.

Price, Michelle. “U.S. Banks Battle Technology Issues in Race for $310 Billion in New Small-Business Aid.” Reuters, April 28, 2020. https://www.reuters.com/article/us-health-coronavirus-banks-lending/u-s-banks-battle-technology- issues-in-race-for-310-billion-in-new-small-business-aid-idUSKCN22A2KO.

FDIC Community Banking Study ■ December 2020 iii PricewaterhouseCoopers. “Consumer Banking and COVID-19 Survey.” July 2020. https://www.pwc.com/us/en/industries/ banking-capital-markets/library/consumer-banking-survey.html.

Sullivan, Richard, and Zhu Wang. “Internet Banking: An Exploration in Technology Diffusion and Impact.” Federal Reserve Bank of Richmond Working Paper no. 13-10, 2013.

Vogels, Emily A. “Millennials Stand Out for Their Technology Use, but Older Generations Also Embrace Digital Life.” Pew Research Center, September 9, 2019. https://www.pewresearch.org/fact-tank/2019/09/09/us-generations-technology-use.

iv FDIC Community Banking Study ■ December 2020 Appendix A: Study Definitions

Summary of FDIC Research Definition of Include: All remaining banking organizations with: Community Banking Organizations — Total assets < indexed size threshold 2 Community banks are designated at the level of the banking organization. — Total assets ≥ indexed size threshold, where:

Loan to assets > 33% (All charters under designated holding companies are • considered community banking charters.) • Core deposits to assets > 50%

Exclude: Any organization with: • More than 1 office but no more than the indexed­ maximum number of offices3 — No loans or no core deposits • Number of large MSAs with offices ≤ 2 — Foreign Assets ≥ 10% of total assets • Number of states with offices ≤ 3 — More than 50% of assets in certain specialty banks, including: • No single office with deposits > indexed maximum branch deposit size.4 • credit card specialists

• consumer nonbank banks1

• industrial loan companies

• trust companies

• bankers’ banks

Lending Specialty Groups Defined for Analysis of FDIC-Insured Community Banks Lending Specialty Group Definition Mortgage Specialists Holds residential mortgage loans greater than 30 percent of total assets Consumer Specialists Holds credit card lines and other loans to individuals greater than 20 percent of total assets Holds construction and development (C&D) loans greater than 10 percent of assets OR total CRE loans CRE Specialists (C&D, multifamily, and secured by other commercial properties) greater than 30 percent of total assets C&I Specialists Holds C&I loans greater than 20 percent of total assets Holds agricultural production loans plus loans secured by farm real estate greater than 20 percent of Agricultural Specialists total assets Meets more than one of the single-specialty definitions above OR holds either retail loans or commercial Multi-Specialists loans greater than 40 percent of total assets No Specialty All other institutions Source: FDIC . Note: All specialty groups require the institution to hold loans greater than 33 percent of total assets .

2 Asset size threshold indexed to equal $250 million in 1985 and $1 65 billion. in 2019 . 3 Maximum number of offices indexed to equal 40 in 1985 and 94 in 1 Consumer nonbank banks are financial institutions with limited 2019 . charters that can make commercial loans or take deposits, but 4 Maximum branch deposit size indexed to equal $1 .25 billion in 1985 not both . and $8 .24 billion in 2019 .

FDIC Community Banking Study ■ December 2020 A-1

Appendix B: Selected Federal Agency Actions Affecting Community Banks, 2008–2019

The federal agency actions listed in this appendix were groupings are for expositional purposes only, and do not compiled on a best efforts basis from the websites of the have any official significance. The groupings are: Federal Deposit Insurance Corporation, Board of Governors • deposit insurance and other federal financial dealings of the Federal Reserve System, Office of the Comptroller of with banks; the Currency, Consumer Financial Protection Bureau, and the Department of the Treasury, including the Financial • capital adequacy; Crimes Enforcement Network, and are intended as a high- • residential mortgage lending and servicing, including level summary of actions by federal regulatory agencies, Home Mortgage Disclosure Act requirements; taken from late December 2007 through year-end 2019, consumer credit and retail payments; that affected community banks. Changes to Call Reports • are excluded, except for one listing pertaining to the • general safety-and-soundness; new FFIEC 051 Call Report. Also excluded is supervisory • Bank Secrecy Act and law enforcement; guidance, except for a model Privacy Act notice and a few Interagency Questions and Answers about flood insurance • bank failure resolution; and the Community Reinvestment Act. Still other excluded • pricing of bank products and services; categories include Statements of Policy, tax and accounting competition and banking industry structure; changes, changes in state law or regulation, inflation • adjustments, actions affecting only internal agency • financial reporting and auditing; procedures, rules relating to the transfer of authority from • other agency actions related to consumers and one agency to another, and rules applying only to large communities; and or internationally active banks. Where a rule is issued by multiple agencies separately, or in both interim-final and • back-office functions. final form, only one listing is included. Links are to the Rule summaries in this table, which may paraphrase or quote announcing press release where available, or to Federal directly from announcing press releases or Federal Register Register notices where a press release is not available. notices without attribution, are deemed accurate but are not Agency actions are grouped by broad topic area; these intended to be relied upon for legal or regulatory purposes.

Deposit Insurance and Other Federal Financial Dealings with Banks The Federal Reserve announced the availability of the Term Auction Facility, a program whereby the Federal December 12, 2007 Reserve would provide term credit to banking organizations against a wider range of collateral than was accepted at its Discount Window (Press Release) . The FDIC issued a rule simplifying the insurance coverage of revocable trust accounts by eliminating the concept September 26, 2008 of “qualifying” beneficiaries and allowing for coverage of virtually any named beneficiary (Press Release) . The Federal Reserve announced that it would begin to pay interest on depository institutions’ required and October 6, 2008 excess reserve balances at the Federal Reserve Banks (Press Release) . The FDIC announced a temporary increase in the standard maximum deposit insurance amount from $100,000 October 7, 2008 to $250,000 pursuant to legislation (this change would be made permanent by law in July 2010) (Press Release) . The FDIC implemented the Temporary Liquidity Guarantee Program (TLGP) to guarantee, for a fee, certain bank October 14, 2008 and holding company obligations, and to implement temporary, unlimited deposit insurance coverage for noninterest-bearing transaction accounts of participating institutions (Press Release) . The Treasury announced the availability of its Capital Purchase Program, under which Treasury would purchase October 14, 2008 up to $250 billion of senior preferred shares of banking organizations, on standardized terms and subject to restrictions on executive compensation and other matters (Press Release) . The FDIC issued a rule governing the payment of deposit insurance assessment dividends when the DIF December 2, 2008 exceeded 1 .35 percent of insured deposits (Federal Register Notice) . December 16, 2008 The FDIC issued a rule that increased deposit insurance assessments uniformly by 7 basis points (Press Release) . continued on page B-2

FDIC Community Banking Study ■ December 2020 B-1 continued from page B-1 The FDIC issued a rule that modified aspects of its risk-based deposit insurance assessment system and March 2, 2009 announced in an interim rule a special deposit insurance assessment (Financial Institution Letter) . The FDIC finalized the special deposit insurance assessment in modified form, setting the assessment at 5 basis May 22, 2009 points on assets minus tier 1 capital (Press Release) . The FDIC issued a rule requiring insured institutions to prepay three years of deposit insurance assessments November 12, 2009 (Press Release) . The FDIC implemented a statutory requirement to provide temporary unlimited deposit insurance coverage of November 9, 2010 noninterest-bearing transaction accounts through the end of 2012 (Press Release) . The FDIC issued a rule setting its designated reserve ratio at 2 percent of estimated insured deposits December 14, 2010 (Press Release) . The FDIC issued a rule that, among other things, implemented a statutory requirement to change the definition February 7, 2011 of the assessment base from adjusted domestic deposits to average consolidated total assets minus average tangible equity (Press Release) . The FDIC issued a rule conforming certain definitions in its assessments regulations to terms used in the Basel III November 24, 2014 revised capital framework (Financial Institution Letter) . The Federal Reserve implemented a rule amending Regulation D (Reserve Requirements of Depository June 18, 2015 Institutions) to make changes to the calculation of interest payments on excess balances maintained by depository institutions at Federal Reserve Banks (Press Release) . The Federal Reserve implemented a statutory requirement by reducing the dividend paid to large banks (with assets greater than $10 billion) on their Federal Reserve bank stock from 6 percent, to the lesser of 6 percent or February 18, 2016 the most recent ten-year Treasury auction rate prior to the dividend, while smaller banks’ dividend rate remained at 6 percent (Press Release) . The FDIC issued a rule to establish a surcharge of 4 .5 cents per $100 of the assessment base on insured institutions with assets greater than $10 billion, implementing a statutory requirement that the assessment cost of increasing the insurance fund from 1 .15 percent of insured deposits to its required level of 1 .35 percent of March 15, 2016 insured deposits should be borne by large institutions rather than by the vast majority of community banks that have assets less than $10 billion, and providing assessment credits to insured institutions of less than $10 billion for the portion of their regular assessments that contribute to growth in the reserve ratio from 1 .15 percent to 1 .35 percent (Press Release) . The FDIC issued a rule revising the methodology used to determine risk-based assessment rates for small banks April 26, 2016 to better differentiate risk (Press Release) . The FDIC issued a rule that made minor technical changes to its assessments regulation April 5, 2018 (Federal Register Notice) . The FDIC issued a rule to allow for alternatives to signature cards for establishing the deposit insurance coverage July 16, 2019 of joint accounts (Press Release) . The FDIC issued a rule providing that small bank assessment credits would be applied when the DIF exceeds 1 .35 November 27, 2019 percent of insured deposits instead of 1 .38 percent of insured deposits (Federal Register Notice) . The FDIC issued a rule that made conforming changes to its assessments regulation to accommodate the December 6, 2019 community bank leverage ratio framework (Federal Register Notice) . Capital Adequacy The Federal Reserve issued a rule permitting bank holding companies to include without limit in tier 1 capital October 16, 2008 senior perpetual preferred stock issued to the Treasury Department (Press Release) . The federal banking agencies announced they would allow banks to treat losses on Fannie Mae and Freddie Mac October 17, 2008 preferred stock as ordinary losses rather than capital losses for regulatory capital purposes, as if a tax change of October 3, 2008, had been enacted in the third quarter (Press Release) . The federal banking agencies issued a rule reducing the amount of the regulatory capital deduction of goodwill December 16, 2008 by the amount of deferred tax liabilities to reflect the maximum exposure to loss in the event of a write-down of goodwill (Press Release) . The Federal Reserve issued a rule extending until 2011 the period of time in which BHCs may include cumulative March 17, 2009 perpetual preferred stock and trust preferred securities in tier 1 capital up to 25 percent of total core capital elements (Press Release) . continued on page B-3

B-2 FDIC Community Banking Study ■ December 2020 continued from page B-2 The Federal Reserve issued rules indicating that senior perpetual preferred stock issued by bank holding companies to the Treasury would count as tier 1 capital, and that subordinated debt issued by S-corps and May 22, 2009 mutual bank holding companies to the Treasury would not count as debt for purposes of the Small Bank Holding Company Policy Statement (Press Release) . The federal banking agencies issued a rule providing that mortgage loans modified under the Making Homes June 26, 2009 Affordable Program would generally have the same risk weight as they had before modification (Press Release) . The federal banking agencies issued a rule defining the risk-based capital treatment of exposures brought onto January 21, 2010 bank balance sheets as a result of Financial Accounting Standards No . 166 and 167 (Press Release) . The Federal Reserve issued a rule that allows small bank holding companies that are S-Corps or that are organized in mutual form to exclude subordinated debt issued to Treasury under the Small Business Lending June 13, 2011 Fund (SBLF) from treatment as “debt” for purposes of the debt-to-equity standard under the Federal Reserve Board’s Small Bank Holding Company Policy Statement (Press Release) . The federal banking agencies issued rules implementing aspects of the Basel III risk-based capital framework . Among other things, the new rules increased by 2 percentage points the agencies’ tier 1 risk-based capital Prompt Corrective Action thresholds defining adequately capitalized and well capitalized banks, introduced a new “common equity tier 1” risk-based capital requirement, tightened the definition of regulatory capital by limiting inclusion of mortgage servicing rights, deferred tax assets and investments in the capital instruments of other financial institutions, excluded future issuances of trust-preferred securities (TruPS) from the tier 1 capital July 2, 2013 of bank holding companies while grandfathering the tier 1 capital treatment of existing TruPS for bank holding companies with assets less than $15 billion, established a new risk-based capital treatment of securitizations that does not rely on credit ratings (implementing a statutory requirement to eliminate references to credit ratings), and changed selected risk weights, including establishing a 250 percent risk weight for amounts of mortgage servicing rights, deferred tax assets, and investments in the capital instruments of other financial institutions that were not deducted from tier 1 capital . For most banks (other than advanced approaches banks), the new rules were effective January 1, 2015 (Press Release) . The federal banking agencies issued rules establishing that banks’ holdings of TruPS as investments were not January 14, 2014 prohibited by the Volcker Rule when those TruPS were those intended to be grandfathered under the Basel III rule (Press Release) . The Federal Reserve issued a rule implementing a statutory requirement to increase the asset threshold used in determining eligibility under its Small Bank Holding Company Policy Statement (SBHCPS) from $500 million to April 9, 2015 $1 billion . BHCs subject to the SBHCPS are not subject to leverage requirements or risk-based capital requirements at the consolidated BHC level . The rule also expanded the applicability of the policy statement to savings and loan holding companies (Press Release) . The federal banking agencies issued rules delaying from taking effect the fully phased-in Basel III deductions for November 21, 2017 mortgage servicing rights, deferred tax assets and investments in the capital instruments of other financial institutions (those deductions had been subject to a multi-year phase-in starting in 2015) (Press Release) . The Federal Reserve issued a rule implementing a statutory requirement to increase the asset threshold used in determining eligibility under its Small Bank Holding Company Policy Statement from $1 billion to $3 billion, August 28, 2018 thereby exempting most BHCs in this size class from being subject to leverage requirements or risk-based capital requirements at the consolidated BHC level (Press Release) . The federal banking agencies issued rules permitting banking organizations the option to phase in over three December 21, 2018 years the day-one impact on regulatory capital of implementing the new Current Expected Credit Loss accounting standard (Press Release) . The federal banking agencies issued rules increasing (for banks not subject to the advanced approaches) the amounts of mortgage servicing rights, deferred tax assets, and investments in the capital instruments of other July 9, 2019 financial institutions that are includable in tier 1 capital . Under the rule, each of these types of exposures can constitute up to 25 percent of tier 1 capital (rather than the previous 10 percent limit), and the previous 15 percent combined limit on the sum of the three types of exposures was eliminated (Press Release) . The federal banking agencies issued rules implementing the statutorily mandated option for qualifying banks with assets less than $10 billion to adopt a Community Bank Leverage Ratio (CBLR) framework . The rule set the October 29, 2019 CBLR at 9 percent . Banks that elect this option will not be subject to risk-based capital requirements unless their tier 1 leverage ratios fall below 9 percent for a period of time (Press Release) . The federal banking agencies issued rules finalizing the risk-based capital treatment of High Volatility Commercial Real Estate (HVCRE) exposures as required by statute . The rules also clarified the risk-based capital November 19, 2019 treatment of land development loans to facilitate the construction of 1–4 family dwellings . Under the rule, such loans would be considered HVCRE and receive a 150 percent risk-weight (Press Release) . continued on page B-4

FDIC Community Banking Study ■ December 2020 B-3 continued from page B-3 Residential Mortgage Lending and Servicing, Including Home Mortgage Disclosure Act Requirements The Federal Reserve issued a rule applying to a newly defined category of “higher-priced mortgage loans .” The rule prohibits a creditor from making a loan without regard to borrowers’ ability to repay the loan from income and assets other than the home’s value, in part based on an analysis of repayment ability based on the highest scheduled payment in the first seven years of the loan; requires creditors to verify borrowers’ income and assets; bans prepayment penalties if the payment can change in the initial four years and otherwise provides that a prepayment penalty period cannot last for more than two years; and requires creditors to establish escrow July 14, 2008 accounts for property taxes and homeowner’s insurance for all first-lien mortgage loans . For any residential mortgage, regardless of whether the loan is higher-priced, creditors may not coerce a real estate appraiser to misstate a home’s value; must provide a good faith estimate of loan costs within three days after loan application, for all mortgages and not just purchase mortgages, and may not charge fees prior to such disclosures (except a reasonable fee for obtaining a credit history); and servicers must not pyramid late fees, must credit payments on receipt and must provide a payoff statement on request (Press Release) . The Federal Reserve revised the rules for reporting price information on higher priced mortgages under October 20, 2008 Regulation C to be consistent with its July 2008 rule . Spreads and thresholds will be based on a survey-based estimate of APRs on comparable mortgages rather than comparable Treasury yields (Press Release) . The Federal Reserve issued a rule implementing a statutory requirement for a seven-day waiting period between a customer’s receipt of required disclosures and the loan closing, and an additional three-day wait if the APR May 8, 2009 changes outside of certain tolerances after the initial disclosure, with the customer having a right to expedite these timelines in case of personal financial emergency (Press Release) . Pursuant to a statutory requirement, the Federal Reserve issued a rule requiring that notice be given to November 16, 2009 borrowers when their mortgage loan has been sold or transferred (Press Release) . Six federal agencies issued a rule implementing statutory requirements for the registration of mortgage loan originators . The rule requires residential mortgage loan originators who are employees of agency-regulated institutions to be registered with the Nationwide Mortgage Licensing System and Registry (registry) . The registry is a database created by the Conference of State Bank Supervisors and the American Association of Residential Mortgage Regulators to support the licensing of mortgage loan originators by the states . Residential mortgage July 28, 2010 loan originators must furnish to the registry information and fingerprints for background checks . The statute generally prohibits employees of agency-regulated institutions from originating residential mortgage loans unless they register with the registry . Each originator will have a unique identifier that will enable consumers to access employment and other background information about that originator from the registry . Under the rule, registered mortgage loan originators and agency-regulated institutions must provide these unique identifiers to consumers (Press Release) . The Federal Reserve issued a rule that applies to mortgage brokers, the companies that employ them, and mortgage loan officers employed by depository institutions and other lenders . The rule prohibits loan originator compensation based on the interest rate or other loan terms, other than the amount of the loan (this is intended to prevent loan originators from increasing their own compensation by raising the consumers’ loan costs); August 16, 2010 prohibits a loan originator that receives compensation from the consumer from also receiving compensation from the lender or another party; and prohibits loan originators from directing or “steering” a consumer to accept a mortgage loan that is not in the consumer’s interest in order to increase the originator’s compensation (Press Release) . The Federal Reserve issued a rule pursuant to statutory requirements that require lenders to disclose how borrowers’ regular mortgage payments can change over time . The rule requires lenders’ disclosures to include a table displaying the initial interest rate and monthly payment; the maximum rate and payment possible in the August 16, 2010 first five years; a worst case example of the maximum rate and payment over the life of the loan; and the fact that consumers might not be able to avoid increased payments by refinancing their loans . The rule also requires lenders to disclose features such as balloon payments or options to make only minimum payments that will cause loan amounts to increase (Press Release) . The Federal Reserve amended its Truth in Lending regulations to ensure that real estate appraisals used in assigning home values are based on the appraiser’s independent professional judgment and that creditors and October 18, 2010 their agents pay customary and reasonable fees to appraisers, implementing a statutory requirement (Press Release) . The Federal Reserve issued a rule clarifying certain disclosure requirements associated with an earlier interim December 22, 2010 final rule (Press Release) . Six federal agencies announced that the Nationwide Mortgage Licensing System and Registry would begin accepting federal registrations . The announcement noted that the rules include an exception for mortgage loan January 31, 2011 originators that originated five or fewer mortgage loans during the previous 12 months and who have never been registered; those loan originators would not be required to complete the federal registration process (Press Release) . continued on page B-5

B-4 FDIC Community Banking Study ■ December 2020 continued from page B-4 The Federal Reserve issued a rule that implemented a statutory provision requiring escrow on jumbo first liens if February 23, 2011 the annual percentage rate (APR) is 2 .5 percentage points or more above the average prime offer rate, rather than the former threshold of 1 .5 percentage points established in the Federal Reserve’s July 2008 rule (Press Release) . The Consumer Financial Protection Bureau (CFPB) issued a rule implementing statutorily required ability-to- repay (ATR) requirements for all new residential mortgages along with certain safe harbors . In general, lenders must document a borrower’s employment status; income and assets; current debt obligations; credit history; monthly payments on the mortgage; monthly payments on any other mortgages on the same property; and monthly payments for mortgage-related obligations . Lenders must evaluate and conclude that the borrower can repay the loan, not just based on introductory or teaser rates but over the life of the loan . Lenders will be presumed to have complied with the ATR rule if they issue “Qualified Mortgages .” These mortgages limit points January 10, 2013 and fees; do not exceed 30 years and do not have interest-only or negative amortization features; and generally will have borrower debt-to-income ratios less than or equal to 43 percent . The rule stated that for a temporary period, loans that do not have a 43 percent debt-to-income ratio but meet government affordability or other standards—such as that they are eligible for purchase by the Federal National Mortgage Association (Fannie Mae) or the Federal Home Loan Mortgage Corporation (Freddie Mac)—will be considered Qualified Mortgages . For higher-priced qualified mortgages, borrowers can rebut the presumption that they had the ability to repay the loan by establishing that they did not have this ability . For lower-priced qualified mortgages, borrowers can only challenge whether the loan met the definition of a qualified mortgage (Press Release) . The CFPB issued a rule applying to high-cost mortgages . For these mortgages, the rule generally bans balloon payments except for certain types of loans made by creditors serving rural or underserved areas, and bans penalties for paying the loan early; bans fees for modifying loans, caps late fees at four percent of the payment that is past due, generally prohibits closing costs from being rolled into the loan amount, and restricts the charging of fees when consumers ask for a payoff statement; prohibits encouraging a consumer to default on an January 10, 2013 existing loan to be refinanced by a high-cost mortgage; and requires consumers to receive housing counseling before taking out a high-cost mortgage . The rule also implements a statutory provision that generally extends the required duration of an escrow account on high-priced mortgage loans from a minimum of one year to a minimum of five years, except for some loans made by creditors that operate predominantly in rural or underserved areas (Press Release) . The CFPB issued a rule addressing mortgage servicing . The rule prohibits servicers from starting a foreclosure proceeding if a borrower has already submitted a complete application for a loan modification or other alternative to foreclosure and the application is still pending review . Servicers cannot make the first notice or filing required for the foreclosure process until a mortgage loan account is more than 120 days delinquent . Servicers must let borrowers know about their “loss mitigation options” to retain their home after borrowers have missed two consecutive payments . Servicers must provide delinquent borrowers with access to employees responsible for helping them . These personnel are responsible for alerting borrowers to any missing information on their applications, telling borrowers about the status of any loss mitigation application, and making sure documents get to the right servicing personnel for processing . The servicer must consider all foreclosure alternatives available from the mortgage owners or investors to help the borrower retain the home . Servicers cannot steer borrowers to those options that are most financially favorable for the servicer . Servicers must consider and respond to a borrower’s application for a loan modification if it arrives at least 37 days before a scheduled foreclosure sale . If the servicer offers an alternative to foreclosure, it must give the borrower time to accept the offer before moving for foreclosure judgment or conducting a foreclosure sale . Servicers cannot foreclose on a property if the borrower and servicer have come to a loss mitigation agreement, unless the borrower fails to perform under that agreement . Servicers must provide regular statements which include: the amount and due date of the next payment; a breakdown of payments by principal, interest, fees, and escrow; and recent transaction activity . January 17, 2013 Servicers must provide a disclosure before the first time the interest rate adjusts for most adjustable-rate mortgages and must provide disclosures before interest rate adjustments that result in a different payment amount . Servicers must have a reasonable basis for concluding that a borrower lacks property insurance before purchasing a new policy . If servicers buy the insurance but receive evidence that it was not needed, they must terminate it within 15 days and refund the premiums . Servicers must credit a consumer’s account the date a payment is received and must credit partial payments in a “suspense account” to the borrowers account once the amount in such an account equals a full payment . Servicers must generally provide a response to consumer requests for the payoff balances of their mortgage loans within seven business days of receiving a written request . Servicers must generally acknowledge receipt of written notices from consumers regarding certain errors or requesting information about their mortgage loans . Generally, within 30 days, the servicer must: correct the error and provide the information requested; conduct a reasonable investigation and inform the borrower why the error did not occur; or inform the borrower that the information requested is unavailable . Servicers must store borrower information in a way that allows it to be easily accessible . Servicers must have policies and procedures in place to ensure that they can provide timely and accurate information to borrowers, investors, and in any foreclosure proceeding, the courts . The rule makes certain exemptions for small servicers that service 5,000 or fewer mortgage loans that they or an affiliate either own or originated . These small servicers are mostly community banks and credit unions servicing mortgages for their customers or members (Press Release) . continued on page B-6

FDIC Community Banking Study ■ December 2020 B-5 continued from page B-5 The CFPB issued a rule implementing statutory provisions requiring that lenders give consumers a copy of each appraisal or other home value estimate free of charge (although a lender generally may still charge the consumer a reasonable fee for the cost of conducting the appraisal or other estimate) . The rule also requires January 18, 2013 that creditors inform consumers within three days of receiving an application for a loan of their right to receive a copy of all appraisals . Creditors are required to provide the copies of appraisal reports and other written home- value estimates to consumers promptly, or three days before closing, whichever is earlier . The rule applies to first-lien mortgages (Press Release) . The CFPB issued a rule addressing steering incentives of mortgage loan originators . The rule prohibits compensation that varies with the loan terms; prohibits loan originator compensation by both the consumer and another person such as the creditor; sets qualification standards for loan originators including character, January 18, 2013 fitness, and financial responsibility reviews, criminal background checks, and training to ensure they have the knowledge about the rules governing the types of loans they originate; and generally prohibits mandatory arbitration of disputes related to mortgage loans and the practice of increasing loan amounts to cover credit insurance premiums . The mandatory arbitration provisions would ultimately be overturned (Press Release) . The CFPB and five other federal agencies issued a rule addressing appraisal requirements for higher-priced mortgage loans . The rule requires creditors to use a licensed or certified appraiser who prepares a written appraisal report based on a physical visit of the interior of the property; requires creditors to provide consumers with a free copy of any appraisal report; and if the seller acquired the property for a lower price during the prior six months and the price difference exceeds certain thresholds, the rule requires creditors to obtain a second January 18, 2013 appraisal at no cost to the consumer (this requirement is intended to address fraudulent property flipping by seeking to ensure that the value of the property legitimately increased) . The rule exempts qualified mortgages, temporary bridge loans and construction loans, loans for new manufactured homes, and loans for mobile homes, trailers, and boats that are dwellings . The rule also has exemptions from the second appraisal requirement to facilitate loans in rural areas and other transactions (Press Release) . The CFPB amended its January 2013 ATR rule by exempting certain nonprofit and community-based lenders that work to help low- and moderate-income consumers obtain affordable housing; extending Qualified Mortgage status to certain loans that small creditors (including community banks and credit unions that have less than $2 billion in assets and each year make 500 or fewer first-lien mortgages) hold in their own portfolios even if the consumers’ debt-to-income ratio exceeds 43 percent; providing a two-year transition period during which small May 29, 2013 lenders can make balloon loans under certain conditions and those loans will meet the definition of Qualified Mortgages; allowing small creditors to charge a higher APR for certain first-lien Qualified Mortgages while maintaining a safe harbor for the ATR requirements; and excluding compensation paid by a lender to a loan originator from counting towards the points and fees threshold used for identifying Qualified Mortgages (Press Release) . The CFPB amended its ability to repay and servicing rules . The rule clarifies and amends how several factors can be used to calculate a consumer’s debt-to-income ratio; explains that CFPB servicing rules do not preempt the field of possible mortgage servicing regulation by states; clarifies which serviced mortgage loans will be July 10, 2013 considered in determining whether a servicer qualifies as small; and clarifies the standards that a loan must meet to be a Qualified Mortgage if the creditor is underwriting it based on GSE or agency guidelines (Press Release) . The CFPB issued revisions to some of its January 2013 mortgage rules . Among other things, the rule clarifies what servicer activities are prohibited in the first 120 days of delinquency; outlines procedures for obtaining follow-up information on loss-mitigation applications; makes it easier for servicers to offer short-term forbearance plans for delinquent borrowers who need only temporary relief without going through a full loss- mitigation evaluation process; clarifies best practices for informing borrowers about the address for error September 13, 2013 resolution documents; pending further study, exempts all small creditors, even those that do not operate predominantly in rural or underserved counties, from the ban on high-cost mortgages featuring balloon payments so long as the loans meet certain restrictions; makes it easier for certain small creditors to continue qualifying for an exemption from a requirement to maintain escrows on certain higher-priced mortgage loans; makes clarifications about financing of credit insurance premiums; and clarifies the circumstances under which a loan originator’s or creditor’s administrative staff acts as loan originators (Press Release) . The CFPB issued a rule requiring new mortgage disclosure forms that replaced then-existing federal disclosures; establishing when the new forms are to be given to the consumer; and limiting how the final deal can change from the original loan estimate . Under the rule, consumers will receive a Loan Estimate within three business November 20, 2013 days after they submit a loan application, replacing the early Truth in Lending statement and the Good Faith Estimate; and they will receive a Closing Disclosure, replacing the final Truth in Lending statement and the HUD-1 settlement statement, three business days before closing (Press Release) . continued on page B-7

B-6 FDIC Community Banking Study ■ December 2020 continued from page B-6 The CFPB and five other federal agencies issued a rule exempting a subset of higher-priced mortgage loans from appraisal requirements . Under the rule, loans of $25,000 or less and certain “streamlined” refinancings are exempt from the appraisal requirements; loans secured by an existing manufactured home and land will be December 12, 2013 subject to the appraisal requirements; loans secured by a new manufactured home and land will be exempt only from the requirement that the appraiser visit the home’s interior; and for loans secured by manufactured homes without land, creditors will be allowed to use other valuation methods without an appraisal (Press Release) . The CFPB issued a rule clarifying that adding the name of an heir to the mortgage of a deceased borrower does July 8, 2014 not trigger the ability-to-repay requirements (Press Release) . The CFPB finalized a rule that helped some nonprofit organizations meet the servicing rule’s requirements for the small servicer exemption; that helped some nonprofit organizations continue to extend certain interest-free, forgivable loans, also known as “soft seconds,” without regard to the 200-mortgage loan limit in the rule while October 22, 2014 still retaining their exemption from the rule; and that clarified the circumstances in which, through January 10, 2021, a lender can refund points and fees after the loan has closed so as to avoid exceeding the cap on points and fees for a Qualified Mortgage (Press Release) . The CFPB issued a rule that extended the deadline within which creditors are required to provide a revised Loan Estimate to within three business days after a consumer locks in a floating interest rate, rather than on the same January 20, 2015 day as required in the original rule . The rule also created a space on the Loan Estimate form where creditors could include language informing consumers that they may receive a revised Loan Estimate for a construction loan that is expected to take more than 60 days to settle (Press Release) . The CFPB and five federal agencies issued a rule implementing statutory requirements to develop standards for April 30, 2015 appraisal management companies . The rule primarily affects state supervision of these companies and a very small number of banks that own or control such companies (Press Release) . The CFPB issued amendments to some of its mortgage rules to, among other things, expand the origination test in the definition of small creditor to creditors originating 2,000 or fewer first lien mortgages per year rather than 500 and exclude loans held in portfolio by the creditor and its affiliates, clarify that the $2 billion asset test September 21, 2015 includes the assets of mortgage-originating affiliates, expand the definition of rural areas and provide a look-up tool to help creditors identify whether a location is rural, and provide additional time (until April 1, 2016) for small creditors’ balloon loans to be considered Qualifying Mortgages (Press Release) . The CFPB issued a rule implementing statutory changes to data collection under the Home Mortgage Disclosure Act (HMDA) . New items required to be reported by covered institutions include the property value, term of the loan, and the duration of any teaser or introductory interest rates; and information about mortgage loan underwriting and pricing, such as an applicant’s debt-to-income ratio, the interest rate of the loan, and the discount points charged for the loan . The rule also requires that covered lenders report, with some exceptions, October 15, 2015 information about all applications and loans secured by dwellings, including reverse mortgages and open-end lines of credit . Small depository institutions located outside a metropolitan statistical area remain excluded from coverage, and in addition, under the rule small depository institutions that have a low loan volume (less than 25 closed-end loans and less than 100 open-end loans over each of the two preceding calendar years) will no longer have to report HMDA data (Press Release) . The CFPB issued a rule implementing a statutory provision that provides broader eligibility for lenders serving rural or underserved areas to originate balloon-payment qualified and high-cost mortgages . Under the rule, a small creditor will be eligible for balloon payment and high-cost balloon payment exemptions from the March 22, 2016 Qualifying Mortgage rule and will not be required to collect escrow for those loans if it originates at least one covered mortgage loan on a property located in a rural or underserved area in the prior calendar year (Press Release) . The CFPB issued a rule that, among other things, requires mortgage servicers to provide certain borrowers with foreclosure protections more than once over the life of the loan; expands consumer protections to surviving family members upon the death of a borrower; requires servicers to provide borrowers in bankruptcy periodic statements with specific information tailored for bankruptcy, as well as a modified written early intervention August 4, 2016 notice to let those borrowers know about loss mitigation options; requires servicers to notify borrowers when loss mitigation applications are complete; clarifies obligations of a new servicer when servicing is transferred; clarifies servicers’ obligations to avoid dual-tracking and prevent wrongful foreclosures; and clarifies when a borrower becomes delinquent (Press Release) . August 24, 2017 The CFPB issued a rule temporarily changing certain HMDA data reporting requirements (Press Release) . The CFPB issued a rule providing greater flexibility and clarity to certain mortgage lenders regarding the September 20, 2017 collection of data about race (Press Release) . The CFPB issued a rule that, among other things, gives servicers a longer, ten-day window to provide required October 4, 2017 early intervention notices to certain consumers at risk of foreclosure who have requested a cease in communication under the Fair Debt Collection Practices Act (Press Release) . continued on page B-8

FDIC Community Banking Study ■ December 2020 B-7 continued from page B-7 The CFPB issued a rule providing a number of technical clarifications regarding HMDA data reporting exemptions August 31, 2018 in light of burden-reducing statutory changes (Press Release) . The CFPB issued a rule that, among other things, extended until January 1, 2022, the CFPB’s temporary HMDA reporting threshold, announced in 2017, for reporting open-end lines of credit . Under the rule, financial October 10, 2019 institutions that originated fewer than 500 open-end lines of credit in either of the two preceding calendar years will not need to collect and report data with respect to open-end lines of credit (Press Release) . The CFPB issued a rule clarifying screening and training requirements for financial institutions that employ loan originators with temporary authority . The rule clarifies that the lender is not required to conduct the screening November 15, 2019 and ensure the training of loan originators with temporary authority, but instead may rely on the screening and training performed by the state as part of its review of the individual’s application for a state loan originator license (Press Release) . Consumer Credit and Retail Payments The Federal Reserve issued a rule prohibiting certain credit card practices by placing limits on interest rate increases during the first year or on pre-existing balances, forbidding “two-cycle billing,” requiring that December 18, 2008 consumers receive a reasonable amount of time to make payments, and limiting fees on subprime cards . The Federal Reserve also revised the disclosures credit-card and revolving-credit customers must receive (Press Release) . Seven federal agencies jointly issued a rule establishing duties of entities that furnish information to credit July 2, 2009 reporting agencies, including the duty to investigate disputes in certain instances at a customer’s request (Press Release) . The Federal Reserve issued a rule requiring lenders to provide written notice to credit card customers 45 days before increasing an interest rate or making other significant changes in terms, notifying them of their ability the July 15, 2009 cancel the card before the terms take effect, and specifying statements be mailed at least 21 days before the payment due date (Press Release) . The Federal Reserve issued a rule prohibiting financial institutions from charging customers fees for paying November 12, 2009 overdrafts on automated teller (ATM) transactions and one-time debit card transactions unless the customer opts into, and receives disclosures about, the institution’s overdraft program (Press Release) . The Federal Reserve and Federal Trade Commission (FTC) issued rules requiring creditors to provide consumers with a notice when the creditor provides credit on less favorable terms than it provides credit to other December 22, 2009 customers, based on a credit report . Customers who receive such notices will be able to obtain a free copy of their credit report to check its accuracy . As an alternative, creditors may provide consumers with a free credit score and information about the score (Press Release) . The Federal Reserve amended aspects of its December 2008 credit card rule, and prohibited the issuance of a credit card to a borrower under the age of 21 unless that person has the ability to make the payments or obtains January 12, 2010 the signature of a parent or co-signer with the ability to do so . The rule also requires creditors to obtain a customer’s consent before charging fees for transactions that exceed the credit limit, and prohibits creditors from allocating payments in a way that maximizes interest charges (Press Release) . The Federal Reserve issued a rule placing restrictions on the fees and expiration dates associated with gift cards . The rules are designed to protect customers against unexpected costs and require that terms and conditions be March 23, 2010 clearly stated . Inactivity fees are not permitted unless the customer has not used the card for at least one year, may not be charged more frequently than once per month, and cards may not expire in less than five years after issuance or last use (Press Release) . The Federal Reserve issued a rule making clarifications and technical changes to two of its earlier rules regarding May 28, 2010 overdraft services (Press Release) . The Federal Reserve issued a rule prohibiting late fees on credit cards of more than $25 unless the borrower has been repeatedly late or the lender can demonstrate the fee is justified by the costs the lender incurs; prohibiting June 15, 2010 penalty fees exceeding the dollar amount of the late payment; prohibiting inactivity fees; prohibiting multiple fees based on a single late payment; and requiring lenders to reconsider whether interest rate increases since January 1, 2009, were warranted (Press Release) . The Federal Reserve issued a rule implementing a statutory extension of the effective date of certain required August 11, 2010 gift card disclosures provided several conditions are met (Press Release) . The Federal Reserve issued a rule prohibiting credit card applications from requesting “household income” (but instead individual income as that more specifically reflects the borrower’s ability to pay); stating that waiving interest for a period of time does not exempt lenders from the requirements of the Credit Card Act (Credit Card March 18, 2011 Accountability Responsibility and Disclosure Act of 2009); and stating that fees charged before an account is opened count toward the Credit Card Act’s fee limitations (i .e ., that fees cannot exceed 25 percent of the account’s initial credit limit) (Press Release) . continued on page B-9

B-8 FDIC Community Banking Study ■ December 2020 continued from page B-8 The Federal Reserve issued a rule that implemented a statutory requirement to expand coverage of truth-in- lending rules to all consumer loans of up to $50,000, with future inflation adjustments, up from the earlier March 25, 2011 threshold of $25,000 . As an exception to these thresholds, truth in lending rules continued to apply to student loans and loans secured by real property regardless of amount (Press Release) . The Federal Reserve and FTC issued a rule revising the content requirements for risk-based pricing notices that July 6, 2011 customers must receive if a credit score is used in setting material terms of credit or in taking adverse action . The rule also revised certain model notices lenders can use to satisfy the disclosure requirements (Press Release) . The FDIC issued a rule describing the requirements that must be satisfied for FDIC-supervised banks to enter into retail foreign exchange transactions with customers . Pursuant to a statutory requirement, a financial institution for which there is a federal regulatory agency shall not enter into retail foreign exchange transactions except in July 12, 2011 compliance with rules established by the relevant regulatory agency . The rule required banks wishing to enter in a foreign exchange business to, among other things, obtain the written consent of the FDIC, maintain records, and provide risk disclosure statements to customers (Federal Register Notice) . The CFPB issued a rule implementing a statutory requirement for providers of international remittances to disclose the exchange rates and fees associated with the transactions and to investigate disputes and remedy January 20, 2012 errors . International money transfers were generally excluded from consumer protection regulations prior to the Dodd Frank Act (Press Release) . The CFPB amended its January 2012 remittance rule to exempt remittance providers making fewer than August 7, 2012 100 remittances per year from being subject to the rule (Press Release) . The CFPB issued a rule reversing the provision of the Federal Reserve’s March 2011 credit card rule which March 22, 2013 included fees charged before account opening in the Credit Card Act’s overall cap on fees . As a result of a court injunction blocking the 2011 provision from taking effect, the CFPB rule eliminated it (Press Release) . The CFPB issued a rule modifying the Federal Reserve’s March 2011 credit card rule with respect to applications April 29, 2013 from stay-at-home spouses or partners . The CFPB rule stated that lenders can consider such a spouse’s or partner’s reasonably anticipated income (Press Release) . The CFPB amended its remittance rule by making the disclosure of foreign taxes or fees charged by the receiving institution optional, provided that the remittance provider disclosed that such fees might apply, and by stating April 30, 2013 that the remittance provider is not liable for losses that result from the sender furnishing incorrect information about the recipient (Press Release) . The CFPB amended its remittance rule by extending a temporary statutory exception allowing institutions to estimate third-party fees and exchange rates when providing remittance transfers to their account holders for August 22, 2014 which they cannot determine exact amounts, and making technical and clarifying changes related to error resolution procedures, permissible methods to deliver disclosures, and other matters (Press Release) . The CFPB issued a rule providing that institutions can post privacy notices online instead of mailing them, if, October 20, 2014 among other things, they only share customer data in a way that does not trigger opt-out requirements . Institutions using this option must use model disclosure forms (Press Release) . The CFPB issued a rule temporarily suspending a requirement that each quarter certain credit card issuers send their agreements to the CFPB, which publishes them in a public database on its website . Card issuers’ obligations to post these agreements on their own publicly available websites remained unaffected . The Credit April 15, 2015 Card Act requires that credit card issuers post consumer credit card agreements on their websites as well as submit those agreements to the CFPB . These agreements feature general terms and conditions, pricing, and fee information (Press Release) . The CFPB issued a rule providing protections for prepaid account users . The rule requires financial institutions to limit consumers’ losses when funds are stolen or cards are lost and to investigate and resolve errors and give October 5, 2016 consumers free access to account information; requires “Know Before You Owe” disclosures about fees and terms for prepaid accounts; and provides protections similar to those for credit cards if consumers are allowed to use credit on their accounts (Press Release) . The CFPB issued a rule applying ability to pay requirements for certain short-term or high-cost loan products such as payday loans, vehicle title loans, deposit advance products, or some longer term balloon loans, and October 5, 2017 restricting lenders’ ability to debit payments on such loans from a borrower’s bank account . Most common types of bank loans were specifically exempted (Press Release) . The CFPB amended its 2016 prepaid accounts rule by, among other things, providing that error resolution and liability limitation protections apply prospectively, after a consumer’s identity has been verified; creating a January 25, 2018 limited exception to the prepaid account rule for certain business relationships involving prepaid accounts linked to traditional credit card products; and allowing negative balances on prepaid accounts in certain circumstances without triggering Regulation Z requirements (Press Release) . continued on page B-10

FDIC Community Banking Study ■ December 2020 B-9 continued from page B-9 The CFPB finalized a rule implementing a legislative provision under which institutions would not have to provide a privacy notice . The conditions are that no opt-out rights are triggered by the institution’s privacy August 10, 2018 policy and no changes have been made to the privacy policy since the most recent disclosure sent to consumers (Press Release) . The CFPB issued a rule implementing statutory amendments to consumers’ rights under the Fair Credit Reporting Act . The rule applies to credit reporting agencies but would affect banks’ ability to access credit September 12, 2018 reports if a customer has requested a freeze on access to these reports based on a possibility of identity theft (Press Release) . General Safety-and-Soundness April 24, 2008 The OCC issued a rule making a number of technical burden-reducing changes to its regulations (Press Release) . The FDIC issued a rule changing its definition of how interest rates substantially exceeding prevailing interest rates would be defined for purposes of implementing the statutory prohibition on banks that are less than well capitalized soliciting and accepting deposits at such “substantially exceeding” interest rates . Under the rule, the May 29, 2009 FDIC would post a national rate for deposits of various types and maturities based on information it received from a data vendor, and a deposit interest rate would “substantially exceed” if it exceeds the corresponding national rate by more than 75 basis points . The earlier definition of substantially exceeds had been based on a comparison to Treasury yields (Press Release) . The OCC issued a rule that incorporated derivatives exposures and securities financing transactions into its legal June 20, 2012 lending limit regulation, as required by statute . The rule included a lookup table approach to limit burden to small institutions (Press Release) . The FDIC issued a rule governing permissible investments of federal and state savings associations, eliminating July 24, 2012 references to credit ratings as required by statute . A similar OCC rule was issued in June of the same year (Federal Register Notice) . The Federal Reserve issued rules implementing the Dodd Frank Act’s requirements for company-run stress tests for banking organizations with consolidated assets exceeding $10 billion . The FDIC and OCC issued substantively October 9, 2012 similar rules . The rules were relevant to large community banks whose actual or planned assets might have exceeded $10 billion as a result of organic growth or merger (Press Release) . Five federal agencies issued rules implementing section 619 of the Dodd Frank Act, also known as the Volcker Rule . The rule prohibited all banking organizations from engaging in proprietary trading as it defined that term, and from owning or sponsoring hedge funds and private equity funds as it defined those terms . On the same December 10, 2013 day, the Federal Reserve announced that banking organizations would have until July 21, 2015, to conform their activities to the new rule . That conformance period subsequently was extended by one year, and again by a second year (Press Release) . Six federal agencies issued a rule implementing statutory risk retention requirements for securitizations . The rule requires securitizers to retain at least 5 percent of the credit risk of securitizations, subject to a number of October 22, 2014 exceptions . While the rule is not relevant to most community banks, a community bank that wished to become an active securitizer would need to determine whether, or how, the rule applies (Press Release) . Five federal agencies issued a rule exempting certain end users of derivatives that are small banks from statutory October 30, 2015 requirements for margin requirements for non-cleared swaps (Press Release) . The federal banking agencies issued a rule expanding the set of institutions eligible for an 18-month examination February 19, 2016 cycle . The maximum asset threshold for eligibility was increased from $500 million to $1 billion, along with other qualifying factors, as a result of a statutory change (Press Release) . The OCC issued a rule that implemented a variety of technical burden-reducing changes to its regulations December 15, 2016 (Press Release) . The OCC issued a rule prohibiting national banks from investing in or dealing in commercial or industrial metals December 28, 2016 (Press Release) . The federal banking agencies issued a rule increasing the threshold for commercial real estate transactions that April 2, 2018 require an appraisal from $250,000 to $500,000 (Press Release) . The federal banking agencies issued a statement explaining, among other things, that company-run stress tests July 6, 2018 would no longer be required for institutions with assets between $10 billion and $100 billion, as a result of the agencies’ implementation of a statutory requirement (Press Release) . The federal banking agencies issued a rule expanding the set of institutions eligible for an 18-month examination August 23, 2018 cycle . The maximum asset threshold for eligibility was increased from $1 billion to $3 billion, along with other qualifying factors, as a result of a statutory change (Press Release) . continued on page B-11

B-10 FDIC Community Banking Study ■ December 2020 continued from page B-10 The FDIC issued a rule implementing a statutory requirement that the FDIC exempt a portion of reciprocal December 19, 2018 deposits from being defined as brokered deposits under certain circumstances (Press Release) . The OCC issued a rule permitting federal savings associations to elect to operate with national bank powers and May 24, 2019 be subject to national bank obligations . The rule implemented a statutory requirement (Press Release) . Five federal agencies issued a rule implementing a statutory exemption of most small banks (banks with July 9, 2019 $10 billion or less in total consolidated assets and total trading assets and liabilities of 5 percent or less of total consolidated assets) from the Volcker Rule (Press Release) . The federal banking agencies issued a rule to increase the threshold for residential real estate transactions September 27, 2019 requiring an appraisal from $250,000 to $400,000 (Press Release) . Bank Secrecy Act and Law Enforcement The Federal Reserve and Treasury issued a rule to implement statutory requirements regarding unlawful internet gambling . The rule requires U .S . financial firms that participate in designated payment systems (including most November 12, 2008 banks) to establish and implement policies and procedures that are reasonably designed to prevent payments to gambling businesses in connection with unlawful internet gambling, provides examples of such policies and procedures, and describes the regulatory enforcement framework (Press Release) . The Financial Crimes Enforcement Network (FinCEN) of the U .S . Treasury issued a rule to simplify the December 4, 2008 requirements for depository institutions to exempt their eligible customers from currency transaction reporting (Press Release) . FinCEN issued a rule that put in place suspicious activity reporting, and customer and transactional information collection requirements on providers and sellers of certain types of prepaid access devices such as plastic cards, July 26, 2011 mobile phones, electronic serial numbers, key fobs, and other mechanisms that provide a portal to funds that have been paid for in advance and are retrievable and transferable . The rule generally exempted small balance products and was issued pursuant to a statutory requirement (Press Release) . The FinCEN and the Federal Reserve announced a rule amending the definitions of “funds transfer” and December 3, 2013 “transmittal of funds” under regulations implementing the Bank Secrecy Act (Press Release) . FinCEN issued a Customer Due Diligence rule requiring financial institutions to identify and verify the identity of the beneficial owners of companies opening accounts; understand the nature and purpose of customer relationships to develop customer risk profiles; and conduct ongoing monitoring to identify and report May 5, 2016 suspicious transactions and, on a risk basis, to maintain and update customer information . With respect to the new requirement to obtain beneficial ownership information, financial institutions will have to identify and verify the identity of any individual who owns 25 percent or more of a legal entity, and an individual who controls the legal entity (Press Release) . Bank Failure Resolution The FDIC issued a rule clarifying how it computes deposit account balances for deposit insurance purposes, and July 17, 2008 requiring institutions to disclose to sweep customers how their sweeps would be treated by the FDIC in the event of the bank’s failure (Financial Institution Letter) . The FDIC issued a rule requiring that, upon written notification from the FDIC, an insured bank in a troubled condition must produce immediately at the close of processing of the institution’s business day, for a period provided in the notification, the electronic files for certain Qualified Financial Contracts’ (QFCs) position and counterparty data; electronic or written lists of QFC counterparty and portfolio location identifiers, certain affiliates of the institution and the institution’s counterparties to QFC transactions, contact information and December 18, 2008 organizational charts for key personnel involved in QFC activities, and contact information for vendors for such activities; and copies of key agreements and related documents for each QFC . The rule allows 60 days from the written notification for an institution to comply and includes provision for additional requests for delay, and includes a de minimis provision such that institutions with fewer than 20 QFC contracts need only have the capability to update records on a daily basis rather than actually provide the records to the FDIC (Financial Institution Letter) . The FDIC issued a rule expanding the QFC recordkeeping requirements (to conform to certain U .S . Treasury regulations) for large insured institutions (assets greater than $50 billion) and, for all other institutions, adding July 31, 2017 and deleting a limited number of QFC data requirements and making certain formatting changes with respect to the QFC recordkeeping requirements (Federal Register Notice) . continued on page B-12

FDIC Community Banking Study ■ December 2020 B-11 continued from page B-11 Pricing of Bank Products and Services The Federal Reserve issued a rule liberalizing the number and type of transfers a customer can make between May 20, 2009 savings and checking accounts, and making it easier for community banks to earn interest on excess balances at Federal Reserve banks (Press Release) . The Federal Reserve issued a rule establishing standards for debit card interchange fees and prohibiting network exclusivity arrangements and routing restrictions, as required by statute . Under the rule, the maximum permissible interchange fee that an issuer may receive for an electronic debit transaction is the sum of 21 cents per transaction and 5 basis points multiplied by the value of the transaction . A related rule allows for an upward adjustment of no more than 1 cent to an issuer’s debit card interchange fee if the issuer develops and implements policies and procedures reasonably designed to achieve certain fraud-prevention standards . If an June 29, 2011 issuer meets these standards and wishes to receive the adjustment, it must certify its eligibility to receive the adjustment to the payment card networks in which it participates . In accordance with the statute, issuers that, together with their affiliates, have assets of less than $10 billion are exempt from the debit card interchange fee standards . The rule prohibits all issuers and networks from restricting the number of networks over which electronic debit transactions may be processed to less than two unaffiliated networks . Issuers and networks are also prohibited from inhibiting a merchant’s ability to direct the routing of the electronic debit transaction over any network that the issuer has enabled to process them (Press Release) . The Federal Reserve issued a rule implementing a statutory requirement to repeal Regulation Q, Prohibition July 14, 2011 Against Payment of Interest on Demand Deposits . The rule was effective July 21, 2011 (Press Release) . Competition and Banking Industry Structure The Federal Reserve issued a rule implementing a statutory prohibition on acquisitions if the resulting company November 5, 2014 has more than 10 percent of all U .S . financial institution liabilities (Press Release) . The federal banking agencies issued a rule increasing the major assets threshold in the management interlocks rule to $10 billion . The major assets prohibition had previously precluded a management official of a depository organization with total assets exceeding $2 .5 billion (or any affiliate of such an organization) from serving at the same time as a management official of an unaffiliated depository organization with total assets exceeding October 2, 2019 $1 .5 billion (or any affiliate of such an organization), regardless of the location of the two depository organizations . Under the rule, the $1 .5 billion and $2 .5 billion thresholds are changed to $10 billion respectively . Other prohibitions in the management interlocks rule, that prevent a management official from serving at the same time as a management official of an unaffiliated depository organization in the same community or relevant metropolitan statistical area, remained unchanged (Press Release) . Financial Reporting and Auditing The FDIC issued a rule applicable to covered insured institutions that, among other things: requires disclosure of the internal control framework and identified material weaknesses; requires management’s assessment of compliance with laws and regulations to disclose any noncompliance; clarifies accountant independence standards; requires certain communications to audit committees; establishes retention requirements for audit working papers; specifies audit committee’s duties regarding the independent public accountant, including ensuring that audit engagement letters do not contain unsafe and unsound limitation of liability provisions; June 23, 2009 requires boards of directors to employ written criteria for evaluating audit committee members’ independence; and states that the assets of a holding company’s bank subsidiaries must be at least 75 percent of the holding company’s consolidated assets for its bank subsidiaries to be able to satisfy the audit requirements at the holding company level . Covered insured institutions are generally those with at least $1 billion in assets for purposes of internal control assessments and at least $500 million for purposes of other requirements (Financial Institution Letter) . The FDIC issued a rule revising its securities disclosure regulations applicable to state nonmember banks with securities required to be registered under section 12 of the Securities Exchange Act of 1934 (Exchange Act) . The November 30, 2010 rule cross-references changes in regulations adopted by the Securities and Exchange Commission (SEC) into the provisions of the FDIC’s securities regulations (Federal Register Notice) . The FDIC issued a rule requiring insured State savings associations and subsidiaries of such State savings associations that act as transfer agents for qualifying securities to register with the FDIC, similar to the May 6, 2016 registration requirements applicable to insured State nonmember banks and subsidiaries of such banks (Federal Register Notice) . The FDIC issued a rule removing certain disclosure requirements applicable to State nonmember banks . The March 20, 2019 disclosures being removed had been made redundant by the availability of more timely and complete information available in Call Reports or on the FDIC’s website (Financial Institution Letter) . continued on page B-13

B-12 FDIC Community Banking Study ■ December 2020 continued from page B-12 The federal banking agencies announced pursuant to a statutory requirement that they would permit insured depository institutions with total assets of less than $5 billion that do not engage in certain complex or June 17, 2019 international activities to file the most streamlined version of the Call Report, the FFIEC 051 Call Report . The previous asset size threshold for use of the FFIEC 051 Call Report was $1 billion . Institutions had begun using the new Call Report as of the March 31, 2017, report date (Press Release) . Other Agency Actions Related to Consumers and Communities August 21, 2008 The OCC issued a rule to encourage public welfare investments by national banks (Press Release) . The federal banking agencies issued revised Interagency Questions and Answers Regarding Community Reinvestment that, among other things, encouraged financial institutions to take steps to help prevent home January 6, 2009 mortgage foreclosures . The agencies use Questions and Answers to assist institutions in compliance with the agencies’ Community Reinvestment Act (CRA) regulations and provide related information to financial institutions and the public (Press Release) . Six federal agencies issued revised Interagency Questions and Answers Regarding Flood Insurance . The July 21, 2009 Questions and Answers, which relate to the agencies’ flood insurance rules, provided technical information on a number of matters (Press Release) . The Federal Reserve issued a rule requiring that private education lenders provide disclosures about loan terms and features at time of application and that they must also disclose information about federal student loan July 30, 2009 programs that may offer less costly alternatives . Additional disclosures are required when the loan is approved and when consummated (Press Release) . Eight federal agencies released a final model privacy notice form intended to make it easier for consumers to understand how financial institutions collect and share information about consumers . Under the Gramm-Leach- November 17, 2009 Bliley Act, institutions must notify consumers of their information-sharing practices and inform consumers of their right to opt out of certain sharing practices . The model form can be used by financial institutions to comply with these requirements (Press Release) . The federal banking agencies issued a rule revising their Community Reinvestment Act regulations to implement September 29, 2010 statutory factors that CRA ratings must consider, including making low-cost higher education loans to low-income borrowers (Press Release) . The federal banking agencies issued rules changing their Community Reinvestment Act regulations to support December 15, 2010 stabilization of communities affected by high foreclosure levels (Press Release) . The federal banking agencies issued revised Interagency Questions and Answers Regarding Community November 15, 2013 Reinvestment that focused on how banks’ support to community development activities may contribute to an outstanding CRA rating (Press Release) . Five federal agencies issued rules implementing statutory flood insurance requirements . The rule requires institutions to escrow flood insurance premiums and fees for loans secured by residential improved real estate or mobile homes made on or after January 1, 2016, unless the loan qualifies for a statutory exception; exempts certain institutions from this escrow requirement if they have total assets of less than $1 billion; requires institutions to provide certain borrowers the option to escrow flood insurance premiums and fees; exempts June 22, 2015 detached structures that are not residences from the requirement to purchase flood insurance (although lenders may choose to require flood insurance); implements statutory provisions regarding force placement by clarifying that regulated lending institutions have the authority to charge a borrower for the cost of force-placed flood insurance coverage beginning on the date on which the borrower’s coverage lapses or becomes insufficient; and identifies when a lender must terminate force-placed flood insurance coverage and refund payments to a borrower (Press Release) . The federal banking agencies issued revised Interagency Questions and Answers Regarding Community July 15, 2016 Reinvestment to assist institutions in compliance with the agencies’ CRA regulations with respect to various matters (Press Release) . The federal banking agencies issued rules amending their respective Community Reinvestment Act regulations to conform to changes made by the CFPB to Regulation C, which implements the Home Mortgage Disclosure Act November 20, 2017 (such consistency has been a practice since 1995, and is intended to make the rules less burdensome), and to eliminate obsolete references to the Neighborhood Stabilization Program (Press Release) . continued on page B-14

FDIC Community Banking Study ■ December 2020 B-13 continued from page B-13 Five federal regulatory agencies issued rules to implement statutory provisions requiring regulated institutions to accept certain private flood insurance policies in addition to National Flood Insurance Program policies . The rule requires that regulated lending institutions accept private flood insurance policies that satisfy criteria February 12, 2019 specified in law; allows institutions to rely on an insurer's written assurances in a private flood insurance policy stating the criteria are met; clarifies that institutions may, under certain conditions, accept private flood insurance policies that do not meet the criteria; and allows institutions to accept certain flood coverage plans provided by mutual aid societies, subject to agency approval (Press Release) . Back-Office Functions The Federal Reserve issued a rule revising its Regulation S, governing the reimbursable costs for financial September 24, 2009 institutions’ providing customer records in response to government agency requests (Press Release) . The Federal Reserve issued a rule revising its Regulation CC . The rule creates a framework for electronic check collection and return and creates new warranties for electronic checks, which will result in a consistent warranty chain regardless of the check’s form . As with existing rules for paper checks, the parties may, by mutual May 31, 2017 agreement, vary the effect of the amendments’ provisions as they apply to electronic checks and electronic returned checks . The final amendments also modify the expeditious-return and notice of nonpayment requirements to create incentives for electronic presentment and return (Press Release) . The FDIC and OCC issued rules to shorten the standard settlement cycle for securities purchased or sold by OCC-supervised and FDIC-supervised institutions . The rule requires banks to settle most securities transactions within the number of business days in the standard settlement cycle followed by registered broker dealers in the United States unless otherwise agreed to by the parties at the time of the transaction . In doing so, the rule aligns June 1, 2018 the settlement cycle requirements of the OCC, FDIC, and Board of Governors of the Federal Reserve System . On September 5, 2017, the securities industry in the United States transitioned from a standard securities settlement cycle of three business days after the date of the contract, commonly known as T+3, to a two-business-day standard, or T+2 (Press Release) . The Federal Reserve issued a rule further amending its Regulation CC . The rule addresses situations where there is a dispute as to whether a check has been altered or was issued with an unauthorized signature, and the September 12, 2018 original paper check is not available for inspection . This rule adopts a presumption of alteration for disputes between banks over whether a substitute check or electronic check contains an alteration or is derived from an original check that was issued with an unauthorized signature of the drawer (Press Release) . The Federal Reserve issued a rule amending its Regulation J, which among other things governs the collection of checks by the Federal Reserve banks and the obligations of parties that send and receive payment items to and from those banks . The amendments clarify and simplify certain provisions of Regulation J, remove obsolete November 15, 2018 provisions, and align the rights and obligations of sending banks, paying banks, and Federal Reserve Banks (Reserve Banks) with the Board’s amendments to Regulation CC to reflect the virtually all-electronic check collection and return environment (Press Release) . The Federal Reserve and the CFPB jointly published amendments to Regulation CC that implement a statutory June 24, 2019 requirement to adjust for inflation the amount of funds depository institutions must make available to their customers (Press Release) .

B-14 FDIC Community Banking Study ■ December 2020 Federal Deposit Insurance Corporation

FDIC-014-2020